Merge pull request #5565 from open-webui/dev

0.3.23

.github/dependabot.yml

@@ -3,10 +3,10 @@ updates:
   - package-ecosystem: pip
     directory: '/backend'
     schedule:
-      interval: weekly
+      interval: monthly
     target-branch: 'dev'
   - package-ecosystem: 'github-actions'
     directory: '/'
     schedule:
       # Check for updates to GitHub Actions every week
-      interval: 'weekly'
+      interval: monthly

.github/workflows/build-release.yml

@@ -10,61 +10,63 @@ jobs:
     runs-on: ubuntu-latest
 
     steps:
-    - name: Checkout repository
-      uses: actions/checkout@v4
+      - name: Checkout repository
+        uses: actions/checkout@v4
 
-    - name: Check for changes in package.json
-      run: |
-        git diff --cached --diff-filter=d package.json || {
-          echo "No changes to package.json"
-          exit 1
-        }
-    
-    - name: Get version number from package.json
-      id: get_version
-      run: |
-        VERSION=$(jq -r '.version' package.json)
-        echo "::set-output name=version::$VERSION"
+      - name: Check for changes in package.json
+        run: |
+          git diff --cached --diff-filter=d package.json || {
+            echo "No changes to package.json"
+            exit 1
+          }
 
-    - name: Extract latest CHANGELOG entry
-      id: changelog
-      run: |
-        CHANGELOG_CONTENT=$(awk 'BEGIN {print_section=0;} /^## \[/ {if (print_section == 0) {print_section=1;} else {exit;}} print_section {print;}' CHANGELOG.md)
-        CHANGELOG_ESCAPED=$(echo "$CHANGELOG_CONTENT" | sed ':a;N;$!ba;s/\n/%0A/g')
-        echo "Extracted latest release notes from CHANGELOG.md:" 
-        echo -e "$CHANGELOG_CONTENT" 
-        echo "::set-output name=content::$CHANGELOG_ESCAPED"
+      - name: Get version number from package.json
+        id: get_version
+        run: |
+          VERSION=$(jq -r '.version' package.json)
+          echo "::set-output name=version::$VERSION"
 
-    - name: Create GitHub release
-      uses: actions/github-script@v7
-      with:
-        github-token: ${{ secrets.GITHUB_TOKEN }}
-        script: |
-          const changelog = `${{ steps.changelog.outputs.content }}`;
-          const release = await github.rest.repos.createRelease({
-            owner: context.repo.owner,
-            repo: context.repo.repo,
-            tag_name: `v${{ steps.get_version.outputs.version }}`,
-            name: `v${{ steps.get_version.outputs.version }}`,
-            body: changelog,
-          })
-          console.log(`Created release ${release.data.html_url}`)
+      - name: Extract latest CHANGELOG entry
+        id: changelog
+        run: |
+          CHANGELOG_CONTENT=$(awk 'BEGIN {print_section=0;} /^## \[/ {if (print_section == 0) {print_section=1;} else {exit;}} print_section {print;}' CHANGELOG.md)
+          CHANGELOG_ESCAPED=$(echo "$CHANGELOG_CONTENT" | sed ':a;N;$!ba;s/\n/%0A/g')
+          echo "Extracted latest release notes from CHANGELOG.md:" 
+          echo -e "$CHANGELOG_CONTENT" 
+          echo "::set-output name=content::$CHANGELOG_ESCAPED"
 
-    - name: Upload package to GitHub release
-      uses: actions/upload-artifact@v4
-      with:
-        name: package
-        path: .
-      env:
-        GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+      - name: Create GitHub release
+        uses: actions/github-script@v7
+        with:
+          github-token: ${{ secrets.GITHUB_TOKEN }}
+          script: |
+            const changelog = `${{ steps.changelog.outputs.content }}`;
+            const release = await github.rest.repos.createRelease({
+              owner: context.repo.owner,
+              repo: context.repo.repo,
+              tag_name: `v${{ steps.get_version.outputs.version }}`,
+              name: `v${{ steps.get_version.outputs.version }}`,
+              body: changelog,
+            })
+            console.log(`Created release ${release.data.html_url}`)
 
-    - name: Trigger Docker build workflow
-      uses: actions/github-script@v7
-      with:
-        script: |
-          github.rest.actions.createWorkflowDispatch({
-            owner: context.repo.owner,
-            repo: context.repo.repo,
-            workflow_id: 'docker-build.yaml',
-            ref: 'v${{ steps.get_version.outputs.version }}',
-          })
+      - name: Upload package to GitHub release
+        uses: actions/upload-artifact@v4
+        with:
+          name: package
+          path: |
+            .
+            !.git
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+
+      - name: Trigger Docker build workflow
+        uses: actions/github-script@v7
+        with:
+          script: |
+            github.rest.actions.createWorkflowDispatch({
+              owner: context.repo.owner,
+              repo: context.repo.repo,
+              workflow_id: 'docker-build.yaml',
+              ref: 'v${{ steps.get_version.outputs.version }}',
+            })

.github/workflows/docker-build.yaml

@@ -312,7 +312,7 @@ jobs:
 
   merge-main-images:
     runs-on: ubuntu-latest
-    needs: [ build-main-image ]
+    needs: [build-main-image]
     steps:
       # GitHub Packages requires the entire repository name to be in lowercase
       # although the repository owner has a lowercase username, this prevents some people from running actions after forking
@@ -364,10 +364,9 @@ jobs:
         run: |
           docker buildx imagetools inspect ${{ env.FULL_IMAGE_NAME }}:${{ steps.meta.outputs.version }}
 
-
   merge-cuda-images:
     runs-on: ubuntu-latest
-    needs: [ build-cuda-image ]
+    needs: [build-cuda-image]
     steps:
       # GitHub Packages requires the entire repository name to be in lowercase
       # although the repository owner has a lowercase username, this prevents some people from running actions after forking
@@ -423,7 +422,7 @@ jobs:
 
   merge-ollama-images:
     runs-on: ubuntu-latest
-    needs: [ build-ollama-image ]
+    needs: [build-ollama-image]
     steps:
       # GitHub Packages requires the entire repository name to be in lowercase
       # although the repository owner has a lowercase username, this prevents some people from running actions after forking

.github/workflows/integration-test.yml

@@ -15,6 +15,13 @@ jobs:
     name: Run Cypress Integration Tests
     runs-on: ubuntu-latest
     steps:
+      - name: Maximize build space
+        uses: AdityaGarg8/remove-unwanted-software@v4.1
+        with:
+          remove-android: 'true'
+          remove-haskell: 'true'
+          remove-codeql: 'true'
+
       - name: Checkout Repository
         uses: actions/checkout@v4
 
@@ -150,7 +157,7 @@ jobs:
           GLOBAL_LOG_LEVEL: debug
         run: |
           cd backend
-          uvicorn main:app --port "8080" --forwarded-allow-ips '*' &
+          uvicorn open_webui.main:app --port "8080" --forwarded-allow-ips '*' &
           UVICORN_PID=$!
           # Wait up to 40 seconds for the server to start
           for i in {1..40}; do
@@ -177,7 +184,7 @@ jobs:
           DATABASE_URL: postgresql://postgres:postgres@localhost:5432/postgres
         run: |
           cd backend
-          uvicorn main:app --port "8081" --forwarded-allow-ips '*' &
+          uvicorn open_webui.main:app --port "8081" --forwarded-allow-ips '*' &
           UVICORN_PID=$!
           # Wait up to 20 seconds for the server to start
           for i in {1..20}; do
@@ -223,7 +230,7 @@ jobs:
 #          DATABASE_URL: mysql://root:mysql@localhost:3306/mysql
 #        run: |
 #          cd backend
-#          uvicorn main:app --port "8083" --forwarded-allow-ips '*' &
+#          uvicorn open_webui.main:app --port "8083" --forwarded-allow-ips '*' &
 #          UVICORN_PID=$!
 #          # Wait up to 20 seconds for the server to start
 #          for i in {1..20}; do

.github/workflows/release-pypi.yml

@@ -4,6 +4,7 @@ on:
   push:
     branches:
       - main # or whatever branch you want to use
+      - pypi-release
 
 jobs:
   release:

CHANGELOG.md

@@ -5,6 +5,239 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [0.3.23] - 2024-09-21
+
+### Added
+
+- **🚀 WebSocket Redis Support**: Enhanced load balancing capabilities for multiple instance setups, promoting better performance and reliability in WebUI.
+- **🔧 Adjustable Chat Controls**: Introduced width-adjustable chat controls, enabling a personalized and more comfortable user interface.
+- **🌎 i18n Updates**: Improved and updated the Chinese translations.
+
+### Fixed
+
+- **🌐 Task Model Unloading Issue**: Modified task handling to use the Ollama /api/chat endpoint instead of OpenAI compatible endpoint, ensuring models stay loaded and ready with custom parameters, thus minimizing delays in task execution.
+- **📝 Title Generation Fix for OpenAI Compatible APIs**: Resolved an issue preventing the generation of titles, enhancing consistency and reliability when using multiple API providers.
+- **🗃️ RAG Duplicate Collection Issue**: Fixed a bug causing repeated processing of the same uploaded file. Now utilizes indexed files to prevent unnecessary duplications, optimizing resource usage.
+- **🖼️ Image Generation Enhancement**: Refactored OpenAI image generation endpoint to be asynchronous, preventing the WebUI from becoming unresponsive during processing, thus enhancing user experience.
+- **🔓 Downgrade Authlib**: Reverted Authlib to version 1.3.1 to address and resolve issues concerning OAuth functionality.
+
+### Changed
+
+- **🔍 Improved Message Interaction**: Enhanced the message node interface to allow for easier focus redirection with a simple click, streamlining user interaction.
+- **✨ Styling Refactor**: Updated WebUI styling for a cleaner, more modern look, enhancing user experience across the platform.
+
+## [0.3.22] - 2024-09-19
+
+### Added
+
+- **⭐ Chat Overview**: Introducing a node-based interactive messages diagram for improved visualization of conversation flows.
+- **🔗 Multiple Vector DB Support**: Now supports multiple vector databases, including the newly added Milvus support. Community contributions for additional database support are highly encouraged!
+- **📡 Experimental Non-Stream Chat Completion**: Experimental feature allowing the use of OpenAI o1 models, which do not support streaming, ensuring more versatile model deployment.
+- **🔍 Experimental Colbert-AI Reranker Integration**: Added support for "jinaai/jina-colbert-v2" as a reranker, enhancing search relevance and accuracy. Note: it may not function at all on low-spec computers.
+- **🕸️ ENABLE_WEBSOCKET_SUPPORT**: Added environment variable for instances to ignore websocket upgrades, stabilizing connections on platforms with websocket issues.
+- **🔊 Azure Speech Service Integration**: Added support for Azure Speech services for Text-to-Speech (TTS).
+- **🎚️ Customizable Playback Speed**: Playback speed control is now available in Call mode settings, allowing users to adjust audio playback speed to their preferences.
+- **🧠 Enhanced Error Messaging**: System now displays helpful error messages directly to users during chat completion issues.
+- **📂 Save Model as Transparent PNG**: Model profile images are now saved as PNGs, supporting transparency and improving visual integration.
+- **📱 iPhone Compatibility Adjustments**: Added padding to accommodate the iPhone navigation bar, improving UI display on these devices.
+- **🔗 Secure Response Headers**: Implemented security response headers, bolstering web application security.
+- **🔧 Enhanced AUTOMATIC1111 Settings**: Users can now configure 'CFG Scale', 'Sampler', and 'Scheduler' parameters directly in the admin settings, enhancing workflow flexibility without source code modifications.
+- **🌍 i18n Updates**: Enhanced translations for Chinese, Ukrainian, Russian, and French, fostering a better localized experience.
+
+### Fixed
+
+- **🛠️ Chat Message Deletion**: Resolved issues with chat message deletion, ensuring a smoother user interaction and system stability.
+- **🔢 Ordered List Numbering**: Fixed the incorrect ordering in lists.
+
+### Changed
+
+- **🎨 Transparent Icon Handling**: Allowed model icons to be displayed on transparent backgrounds, improving UI aesthetics.
+- **📝 Improved RAG Template**: Enhanced Retrieval-Augmented Generation template, optimizing context handling and error checking for more precise operation.
+
+## [0.3.21] - 2024-09-08
+
+### Added
+
+- **📊 Document Count Display**: Now displays the total number of documents directly within the dashboard.
+- **🚀 Ollama Embed API Endpoint**: Enabled /api/embed endpoint proxy support.
+
+### Fixed
+
+- **🐳 Docker Launch Issue**: Resolved the problem preventing Open-WebUI from launching correctly when using Docker.
+
+### Changed
+
+- **🔍 Enhanced Search Prompts**: Improved the search query generation prompts for better accuracy and user interaction, enhancing the overall search experience.
+
+## [0.3.20] - 2024-09-07
+
+### Added
+
+- **🌐 Translation Update**: Updated Catalan translations to improve user experience for Catalan speakers.
+
+### Fixed
+
+- **📄 PDF Download**: Resolved a configuration issue with fonts directory, ensuring PDFs are now downloaded with the correct formatting.
+- **🛠️ Installation of Tools & Functions Requirements**: Fixed a bug where necessary requirements for tools and functions were not properly installing.
+- **🔗 Inline Image Link Rendering**: Enabled rendering of images directly from links in chat.
+- **📞 Post-Call User Interface Cleanup**: Adjusted UI behavior to automatically close chat controls after a voice call ends, reducing screen clutter.
+- **🎙️ Microphone Deactivation Post-Call**: Addressed an issue where the microphone remained active after calls.
+- **✍️ Markdown Spacing Correction**: Corrected spacing in Markdown rendering, ensuring text appears neatly and as expected.
+- **🔄 Message Re-rendering**: Fixed an issue causing all response messages to re-render with each new message, now improving chat performance.
+
+### Changed
+
+- **🌐 Refined Web Search Integration**: Deprecated the Search Query Generation Prompt threshold; introduced a toggle button for "Enable Web Search Query Generation" allowing users to opt-in to using web search more judiciously.
+- **📝 Default Prompt Templates Update**: Emptied environment variable templates for search and title generation now default to the Open WebUI default prompt templates, simplifying configuration efforts.
+
+## [0.3.19] - 2024-09-05
+
+### Added
+
+- **🌐 Translation Update**: Improved Chinese translations.
+
+### Fixed
+
+- **📂 DATA_DIR Overriding**: Fixed an issue to avoid overriding DATA_DIR, preventing errors when directories are set identically, ensuring smoother operation and data management.
+- **🛠️ Frontmatter Extraction**: Fixed the extraction process for frontmatter in tools and functions.
+
+### Changed
+
+- **🎨 UI Styling**: Refined the user interface styling for enhanced visual coherence and user experience.
+
+## [0.3.18] - 2024-09-04
+
+### Added
+
+- **🛠️ Direct Database Execution for Tools & Functions**: Enhanced the execution of Python files for tools and functions, now directly loading from the database for a more streamlined backend process.
+
+### Fixed
+
+- **🔄 Automatic Rewrite of Import Statements in Tools & Functions**: Tool and function scripts that import 'utils', 'apps', 'main', 'config' will now automatically rename these with 'open_webui.', ensuring compatibility and consistency across different modules.
+- **🎨 Styling Adjustments**: Minor fixes in the visual styling to improve user experience and interface consistency.
+
+## [0.3.17] - 2024-09-04
+
+### Added
+
+- **🔄 Import/Export Configuration**: Users can now import and export webui configurations from admin settings > Database, simplifying setup replication across systems.
+- **🌍 Web Search via URL Parameter**: Added support for activating web search directly through URL by setting 'web-search=true'.
+- **🌐 SearchApi Integration**: Added support for SearchApi as an alternative web search provider, enhancing search capabilities within the platform.
+- **🔍 Literal Type Support in Tools**: Tools now support the Literal type.
+- **🌍 Updated Translations**: Improved translations for Chinese, Ukrainian, and Catalan.
+
+### Fixed
+
+- **🔧 Pip Install Issue**: Resolved the issue where pip install failed due to missing 'alembic.ini', ensuring smoother installation processes.
+- **🌃 Automatic Theme Update**: Fixed an issue where the color theme did not update dynamically with system changes.
+- **🛠️ User Agent in ComfyUI**: Added default headers in ComfyUI to fix access issues, improving reliability in network communications.
+- **🔄 Missing Chat Completion Response Headers**: Ensured proper return of proxied response headers during chat completion, improving API reliability.
+- **🔗 Websocket Connection Prioritization**: Modified socket.io configuration to prefer websockets and more reliably fallback to polling, enhancing connection stability.
+- **🎭 Accessibility Enhancements**: Added missing ARIA labels for buttons, improving accessibility for visually impaired users.
+- **⚖️ Advanced Parameter**: Fixed an issue ensuring that advanced parameters are correctly applied in all scenarios, ensuring consistent behavior of user-defined settings.
+
+### Changed
+
+- **🔁 Namespace Reorganization**: Reorganized all Python files under the 'open_webui' namespace to streamline the project structure and improve maintainability. Tools and functions importing from 'utils' should now use 'open_webui.utils'.
+- **🚧 Dependency Updates**: Updated several backend dependencies like 'aiohttp', 'authlib', 'duckduckgo-search', 'flask-cors', and 'langchain' to their latest versions, enhancing performance and security.
+
+## [0.3.16] - 2024-08-27
+
+### Added
+
+- **🚀 Config DB Migration**: Migrated configuration handling from config.json to the database, enabling high-availability setups and load balancing across multiple Open WebUI instances.
+- **🔗 Call Mode Activation via URL**: Added a 'call=true' URL search parameter enabling direct shortcuts to activate call mode, enhancing user interaction on mobile devices.
+- **✨ TTS Content Control**: Added functionality to control how message content is segmented for Text-to-Speech (TTS) generation requests, allowing for more flexible speech output options.
+- **😄 Show Knowledge Search Status**: Enhanced model usage transparency by displaying status when working with knowledge-augmented models, helping users understand the system's state during queries.
+- **👆 Click-to-Copy for Codespan**: Enhanced interactive experience in the WebUI by allowing users to click to copy content from code spans directly.
+- **🚫 API User Blocking via Model Filter**: Introduced the ability to block API users based on customized model filters, enhancing security and control over API access.
+- **🎬 Call Overlay Styling**: Adjusted call overlay styling on large screens to not cover the entire interface, but only the chat control area, for a more unobtrusive interaction experience.
+
+### Fixed
+
+- **🔧 LaTeX Rendering Issue**: Addressed an issue that affected the correct rendering of LaTeX.
+- **📁 File Leak Prevention**: Resolved the issue of uploaded files mistakenly being accessible across user chats.
+- **🔧 Pipe Functions with '**files**' Param**: Fixed issues with '**files**' parameter not functioning correctly in pipe functions.
+- **📝 Markdown Processing for RAG**: Fixed issues with processing Markdown in files.
+- **🚫 Duplicate System Prompts**: Fixed bugs causing system prompts to duplicate.
+
+### Changed
+
+- **🔋 Wakelock Permission**: Optimized the activation of wakelock to only engage during call mode, conserving device resources and improving battery performance during idle periods.
+- **🔍 Content-Type for Ollama Chats**: Added 'application/x-ndjson' content-type to '/api/chat' endpoint responses to match raw Ollama responses.
+- **✋ Disable Signups Conditionally**: Implemented conditional logic to disable sign-ups when 'ENABLE_LOGIN_FORM' is set to false.
+
+## [0.3.15] - 2024-08-21
+
+### Added
+
+- **🔗 Temporary Chat Activation**: Integrated a new URL parameter 'temporary-chat=true' to enable temporary chat sessions directly through the URL.
+- **🌄 ComfyUI Seed Node Support**: Introduced seed node support in ComfyUI for image generation, allowing users to specify node IDs for randomized seed assignment.
+
+### Fixed
+
+- **🛠️ Tools and Functions**: Resolved a critical issue where Tools and Functions were not properly functioning, restoring full capability and reliability to these essential features.
+- **🔘 Chat Action Button in Many Model Chat**: Fixed the malfunctioning of chat action buttons in many model chat environments, ensuring a smoother and more responsive user interaction.
+- **⏪ Many Model Chat Compatibility**: Restored backward compatibility for many model chats.
+
+## [0.3.14] - 2024-08-21
+
+### Added
+
+- **🛠️ Custom ComfyUI Workflow**: Deprecating several older environment variables, this enhancement introduces a new, customizable workflow for a more tailored user experience.
+- **🔀 Merge Responses in Many Model Chat**: Enhances the dialogue by merging responses from multiple models into a single, coherent reply, improving the interaction quality in many model chats.
+- **✅ Multiple Instances of Same Model in Chats**: Enhanced many model chat to support adding multiple instances of the same model.
+- **🔧 Quick Actions in Model Workspace**: Enhanced Shift key quick actions for hiding/unhiding and deleting models, facilitating a smoother workflow.
+- **🗨️ Markdown Rendering in User Messages**: User messages are now rendered in Markdown, enhancing readability and interaction.
+- **💬 Temporary Chat Feature**: Introduced a temporary chat feature, deprecating the old chat history setting to enhance user interaction flexibility.
+- **🖋️ User Message Editing**: Enhanced the user chat editing feature to allow saving changes without sending, providing more flexibility in message management.
+- **🛡️ Security Enhancements**: Various security improvements implemented across the platform to ensure safer user experiences.
+- **🌍 Updated Translations**: Enhanced translations for Chinese, Ukrainian, and Bahasa Malaysia, improving localization and user comprehension.
+
+### Fixed
+
+- **📑 Mermaid Rendering Issue**: Addressed issues with Mermaid chart rendering to ensure clean and clear visual data representation.
+- **🎭 PWA Icon Maskability**: Fixed the Progressive Web App icon to be maskable, ensuring proper display on various device home screens.
+- **🔀 Cloned Model Chat Freezing Issue**: Fixed a bug where cloning many model chats would cause freezing, enhancing stability and responsiveness.
+- **🔍 Generic Error Handling and Refinements**: Various minor fixes and refinements to address previously untracked issues, ensuring smoother operations.
+
+### Changed
+
+- **🖼️ Image Generation Refactor**: Overhauled image generation processes for improved efficiency and quality.
+- **🔨 Refactor Tool and Function Calling**: Refactored tool and function calling mechanisms for improved clarity and maintainability.
+- **🌐 Backend Library Updates**: Updated critical backend libraries including SQLAlchemy, uvicorn[standard], faster-whisper, bcrypt, and boto3 for enhanced performance and security.
+
+### Removed
+
+- **🚫 Deprecated ComfyUI Environment Variables**: Removed several outdated environment variables related to ComfyUI settings, simplifying configuration management.
+
+## [0.3.13] - 2024-08-14
+
+### Added
+
+- **🎨 Enhanced Markdown Rendering**: Significant improvements in rendering markdown, ensuring smooth and reliable display of LaTeX and Mermaid charts, enhancing user experience with more robust visual content.
+- **🔄 Auto-Install Tools & Functions Python Dependencies**: For 'Tools' and 'Functions', Open WebUI now automatically install extra python requirements specified in the frontmatter, streamlining setup processes and customization.
+- **🌀 OAuth Email Claim Customization**: Introduced an 'OAUTH_EMAIL_CLAIM' variable to allow customization of the default "email" claim within OAuth configurations, providing greater flexibility in authentication processes.
+- **📶 Websocket Reconnection**: Enhanced reliability with the capability to automatically reconnect when a websocket is closed, ensuring consistent and stable communication.
+- **🤳 Haptic Feedback on Support Devices**: Android devices now support haptic feedback for an immersive tactile experience during certain interactions.
+
+### Fixed
+
+- **🛠️ ComfyUI Performance Improvement**: Addressed an issue causing FastAPI to stall when ComfyUI image generation was active; now runs in a separate thread to prevent UI unresponsiveness.
+- **🔀 Session Handling**: Fixed an issue mandating session_id on client-side to ensure smoother session management and transitions.
+- **🖋️ Minor Bug Fixes and Format Corrections**: Various minor fixes including typo corrections, backend formatting improvements, and test amendments enhancing overall system stability and performance.
+
+### Changed
+
+- **🚀 Migration to SvelteKit 2**: Upgraded the underlying framework to SvelteKit version 2, offering enhanced speed, better code structure, and improved deployment capabilities.
+- **🧹 General Cleanup and Refactoring**: Performed broad cleanup and refactoring across the platform, improving code efficiency and maintaining high standards of code health.
+- **🚧 Integration Testing Improvements**: Modified how Cypress integration tests detect chat messages and updated sharing tests for better reliability and accuracy.
+- **📁 Standardized '.safetensors' File Extension**: Renamed the '.sft' file extension to '.safetensors' for ComfyUI workflows, standardizing file formats across the platform.
+
+### Removed
+
+- **🗑️ Deprecated Frontend Functions**: Removed frontend functions that were migrated to backend to declutter the codebase and reduce redundancy.
+
 ## [0.3.12] - 2024-08-07
 
 ### Added

Dockerfile

@@ -74,6 +74,10 @@ ENV RAG_EMBEDDING_MODEL="$USE_EMBEDDING_MODEL_DOCKER" \
 
 ## Hugging Face download cache ##
 ENV HF_HOME="/app/backend/data/cache/embedding/models"
+
+## Torch Extensions ##
+# ENV TORCH_EXTENSIONS_DIR="/.cache/torch_extensions"
+
 #### Other models ##########################################################
 
 WORKDIR /app/backend
@@ -96,7 +100,7 @@ RUN chown -R $UID:$GID /app $HOME
 RUN if [ "$USE_OLLAMA" = "true" ]; then \
     apt-get update && \
     # Install pandoc and netcat
-    apt-get install -y --no-install-recommends pandoc netcat-openbsd curl && \
+    apt-get install -y --no-install-recommends git build-essential pandoc netcat-openbsd curl && \
     apt-get install -y --no-install-recommends gcc python3-dev && \
     # for RAG OCR
     apt-get install -y --no-install-recommends ffmpeg libsm6 libxext6 && \
@@ -109,7 +113,7 @@ RUN if [ "$USE_OLLAMA" = "true" ]; then \
     else \
     apt-get update && \
     # Install pandoc, netcat and gcc
-    apt-get install -y --no-install-recommends pandoc gcc netcat-openbsd curl jq && \
+    apt-get install -y --no-install-recommends git build-essential pandoc gcc netcat-openbsd curl jq && \
     apt-get install -y --no-install-recommends gcc python3-dev && \
     # for RAG OCR
     apt-get install -y --no-install-recommends ffmpeg libsm6 libxext6 && \
@@ -157,5 +161,6 @@ USER $UID:$GID
 
 ARG BUILD_HASH
 ENV WEBUI_BUILD_VERSION=${BUILD_HASH}
+ENV DOCKER true
 
 CMD [ "bash", "start.sh"]

README.md

@@ -37,7 +37,7 @@ Open WebUI is an [extensible](https://github.com/open-webui/pipelines), feature-
 
 - 📚 **Local RAG Integration**: Dive into the future of chat interactions with groundbreaking Retrieval Augmented Generation (RAG) support. This feature seamlessly integrates document interactions into your chat experience. You can load documents directly into the chat or add files to your document library, effortlessly accessing them using the `#` command before a query.
 
-- 🔍 **Web Search for RAG**: Perform web searches using providers like `SearXNG`, `Google PSE`, `Brave Search`, `serpstack`, `serper`, `Serply`, `DuckDuckGo` and `TavilySearch` and inject the results directly into your chat experience.
+- 🔍 **Web Search for RAG**: Perform web searches using providers like `SearXNG`, `Google PSE`, `Brave Search`, `serpstack`, `serper`, `Serply`, `DuckDuckGo`, `TavilySearch` and `SearchApi` and inject the results directly into your chat experience.
 
 - 🌐 **Web Browsing Capability**: Seamlessly integrate websites into your chat experience using the `#` command followed by a URL. This feature allows you to incorporate web content directly into your conversations, enhancing the richness and depth of your interactions.
 

backend/.gitignore

@@ -8,9 +8,5 @@ _test
 Pipfile
 !/data
 /data/*
-!/data/litellm
-/data/litellm/*
-!data/litellm/config.yaml
-
-!data/config.json
+/open_webui/data/*
 .webui_secret_key

backend/apps/images/utils/comfyui.py

@@ -1,407 +0,0 @@
-import websocket  # NOTE: websocket-client (https://github.com/websocket-client/websocket-client)
-import uuid
-import json
-import urllib.request
-import urllib.parse
-import random
-import logging
-
-from config import SRC_LOG_LEVELS
-
-log = logging.getLogger(__name__)
-log.setLevel(SRC_LOG_LEVELS["COMFYUI"])
-
-from pydantic import BaseModel
-
-from typing import Optional
-
-COMFYUI_DEFAULT_PROMPT = """
-{
-  "3": {
-    "inputs": {
-      "seed": 0,
-      "steps": 20,
-      "cfg": 8,
-      "sampler_name": "euler",
-      "scheduler": "normal",
-      "denoise": 1,
-      "model": [
-        "4",
-        0
-      ],
-      "positive": [
-        "6",
-        0
-      ],
-      "negative": [
-        "7",
-        0
-      ],
-      "latent_image": [
-        "5",
-        0
-      ]
-    },
-    "class_type": "KSampler",
-    "_meta": {
-      "title": "KSampler"
-    }
-  },
-  "4": {
-    "inputs": {
-      "ckpt_name": "model.safetensors"
-    },
-    "class_type": "CheckpointLoaderSimple",
-    "_meta": {
-      "title": "Load Checkpoint"
-    }
-  },
-  "5": {
-    "inputs": {
-      "width": 512,
-      "height": 512,
-      "batch_size": 1
-    },
-    "class_type": "EmptyLatentImage",
-    "_meta": {
-      "title": "Empty Latent Image"
-    }
-  },
-  "6": {
-    "inputs": {
-      "text": "Prompt",
-      "clip": [
-        "4",
-        1
-      ]
-    },
-    "class_type": "CLIPTextEncode",
-    "_meta": {
-      "title": "CLIP Text Encode (Prompt)"
-    }
-  },
-  "7": {
-    "inputs": {
-      "text": "Negative Prompt",
-      "clip": [
-        "4",
-        1
-      ]
-    },
-    "class_type": "CLIPTextEncode",
-    "_meta": {
-      "title": "CLIP Text Encode (Prompt)"
-    }
-  },
-  "8": {
-    "inputs": {
-      "samples": [
-        "3",
-        0
-      ],
-      "vae": [
-        "4",
-        2
-      ]
-    },
-    "class_type": "VAEDecode",
-    "_meta": {
-      "title": "VAE Decode"
-    }
-  },
-  "9": {
-    "inputs": {
-      "filename_prefix": "ComfyUI",
-      "images": [
-        "8",
-        0
-      ]
-    },
-    "class_type": "SaveImage",
-    "_meta": {
-      "title": "Save Image"
-    }
-  }
-}
-"""
-
-FLUX_DEFAULT_PROMPT = """
-{
-    "5": {
-        "inputs": {
-            "width": 1024,
-            "height": 1024,
-            "batch_size": 1
-        },
-        "class_type": "EmptyLatentImage"
-    },
-    "6": {
-        "inputs": {
-            "text": "Input Text Here",
-            "clip": [
-                "11",
-                0
-            ]
-        },
-        "class_type": "CLIPTextEncode"
-    },
-    "8": {
-        "inputs": {
-            "samples": [
-                "13",
-                0
-            ],
-            "vae": [
-                "10",
-                0
-            ]
-        },
-        "class_type": "VAEDecode"
-    },
-    "9": {
-        "inputs": {
-            "filename_prefix": "ComfyUI",
-            "images": [
-                "8",
-                0
-            ]
-        },
-        "class_type": "SaveImage"
-    },
-    "10": {
-        "inputs": {
-            "vae_name": "ae.sft"
-        },
-        "class_type": "VAELoader"
-    },
-    "11": {
-        "inputs": {
-            "clip_name1": "clip_l.safetensors",
-            "clip_name2": "t5xxl_fp16.safetensors",
-            "type": "flux"
-        },
-        "class_type": "DualCLIPLoader"
-    },
-    "12": {
-        "inputs": {
-            "unet_name": "flux1-dev.sft",
-            "weight_dtype": "default"
-        },
-        "class_type": "UNETLoader"
-    },
-    "13": {
-        "inputs": {
-            "noise": [
-                "25",
-                0
-            ],
-            "guider": [
-                "22",
-                0
-            ],
-            "sampler": [
-                "16",
-                0
-            ],
-            "sigmas": [
-                "17",
-                0
-            ],
-            "latent_image": [
-                "5",
-                0
-            ]
-        },
-        "class_type": "SamplerCustomAdvanced"
-    },
-    "16": {
-        "inputs": {
-            "sampler_name": "euler"
-        },
-        "class_type": "KSamplerSelect"
-    },
-    "17": {
-        "inputs": {
-            "scheduler": "simple",
-            "steps": 20,
-            "denoise": 1,
-            "model": [
-                "12",
-                0
-            ]
-        },
-        "class_type": "BasicScheduler"
-    },
-    "22": {
-        "inputs": {
-            "model": [
-                "12",
-                0
-            ],
-            "conditioning": [
-                "6",
-                0
-            ]
-        },
-        "class_type": "BasicGuider"
-    },
-    "25": {
-        "inputs": {
-            "noise_seed": 778937779713005
-        },
-        "class_type": "RandomNoise"
-    }
-}
-"""
-
-
-def queue_prompt(prompt, client_id, base_url):
-    log.info("queue_prompt")
-    p = {"prompt": prompt, "client_id": client_id}
-    data = json.dumps(p).encode("utf-8")
-    req = urllib.request.Request(f"{base_url}/prompt", data=data)
-    return json.loads(urllib.request.urlopen(req).read())
-
-
-def get_image(filename, subfolder, folder_type, base_url):
-    log.info("get_image")
-    data = {"filename": filename, "subfolder": subfolder, "type": folder_type}
-    url_values = urllib.parse.urlencode(data)
-    with urllib.request.urlopen(f"{base_url}/view?{url_values}") as response:
-        return response.read()
-
-
-def get_image_url(filename, subfolder, folder_type, base_url):
-    log.info("get_image")
-    data = {"filename": filename, "subfolder": subfolder, "type": folder_type}
-    url_values = urllib.parse.urlencode(data)
-    return f"{base_url}/view?{url_values}"
-
-
-def get_history(prompt_id, base_url):
-    log.info("get_history")
-    with urllib.request.urlopen(f"{base_url}/history/{prompt_id}") as response:
-        return json.loads(response.read())
-
-
-def get_images(ws, prompt, client_id, base_url):
-    prompt_id = queue_prompt(prompt, client_id, base_url)["prompt_id"]
-    output_images = []
-    while True:
-        out = ws.recv()
-        if isinstance(out, str):
-            message = json.loads(out)
-            if message["type"] == "executing":
-                data = message["data"]
-                if data["node"] is None and data["prompt_id"] == prompt_id:
-                    break  # Execution is done
-        else:
-            continue  # previews are binary data
-
-    history = get_history(prompt_id, base_url)[prompt_id]
-    for o in history["outputs"]:
-        for node_id in history["outputs"]:
-            node_output = history["outputs"][node_id]
-            if "images" in node_output:
-                for image in node_output["images"]:
-                    url = get_image_url(
-                        image["filename"], image["subfolder"], image["type"], base_url
-                    )
-                    output_images.append({"url": url})
-    return {"data": output_images}
-
-
-class ImageGenerationPayload(BaseModel):
-    prompt: str
-    negative_prompt: Optional[str] = ""
-    steps: Optional[int] = None
-    seed: Optional[int] = None
-    width: int
-    height: int
-    n: int = 1
-    cfg_scale: Optional[float] = None
-    sampler: Optional[str] = None
-    scheduler: Optional[str] = None
-    sd3: Optional[bool] = None
-    flux: Optional[bool] = None
-    flux_weight_dtype: Optional[str] = None
-    flux_fp8_clip: Optional[bool] = None
-
-
-def comfyui_generate_image(
-    model: str, payload: ImageGenerationPayload, client_id, base_url
-):
-    ws_url = base_url.replace("http://", "ws://").replace("https://", "wss://")
-
-    comfyui_prompt = json.loads(COMFYUI_DEFAULT_PROMPT)
-
-    if payload.cfg_scale:
-        comfyui_prompt["3"]["inputs"]["cfg"] = payload.cfg_scale
-
-    if payload.sampler:
-        comfyui_prompt["3"]["inputs"]["sampler"] = payload.sampler
-
-    if payload.scheduler:
-        comfyui_prompt["3"]["inputs"]["scheduler"] = payload.scheduler
-
-    if payload.sd3:
-        comfyui_prompt["5"]["class_type"] = "EmptySD3LatentImage"
-
-    if payload.steps:
-        comfyui_prompt["3"]["inputs"]["steps"] = payload.steps
-
-    comfyui_prompt["4"]["inputs"]["ckpt_name"] = model
-    comfyui_prompt["7"]["inputs"]["text"] = payload.negative_prompt
-    comfyui_prompt["3"]["inputs"]["seed"] = (
-        payload.seed if payload.seed else random.randint(0, 18446744073709551614)
-    )
-
-    # as Flux uses a completely different workflow, we must treat it specially
-    if payload.flux:
-        comfyui_prompt = json.loads(FLUX_DEFAULT_PROMPT)
-        comfyui_prompt["12"]["inputs"]["unet_name"] = model
-        comfyui_prompt["25"]["inputs"]["noise_seed"] = (
-            payload.seed if payload.seed else random.randint(0, 18446744073709551614)
-        )
-
-        if payload.sampler:
-            comfyui_prompt["16"]["inputs"]["sampler_name"] = payload.sampler
-
-        if payload.steps:
-            comfyui_prompt["17"]["inputs"]["steps"] = payload.steps
-
-        if payload.scheduler:
-            comfyui_prompt["17"]["inputs"]["scheduler"] = payload.scheduler
-
-        if payload.flux_weight_dtype:
-            comfyui_prompt["12"]["inputs"]["weight_dtype"] = payload.flux_weight_dtype
-
-        if payload.flux_fp8_clip:
-            comfyui_prompt["11"]["inputs"][
-                "clip_name2"
-            ] = "t5xxl_fp8_e4m3fn.safetensors"
-
-    comfyui_prompt["5"]["inputs"]["batch_size"] = payload.n
-    comfyui_prompt["5"]["inputs"]["width"] = payload.width
-    comfyui_prompt["5"]["inputs"]["height"] = payload.height
-
-    # set the text prompt for our positive CLIPTextEncode
-    comfyui_prompt["6"]["inputs"]["text"] = payload.prompt
-
-    try:
-        ws = websocket.WebSocket()
-        ws.connect(f"{ws_url}/ws?clientId={client_id}")
-        log.info("WebSocket connection established.")
-    except Exception as e:
-        log.exception(f"Failed to connect to WebSocket server: {e}")
-        return None
-
-    try:
-        images = get_images(ws, comfyui_prompt, client_id, base_url)
-    except Exception as e:
-        log.exception(f"Error while receiving images: {e}")
-        images = None
-
-    ws.close()
-
-    return images

backend/apps/webui/utils.py

@@ -1,90 +0,0 @@
-from importlib import util
-import os
-import re
-
-from config import TOOLS_DIR, FUNCTIONS_DIR
-
-
-def extract_frontmatter(file_path):
-    """
-    Extract frontmatter as a dictionary from the specified file path.
-    """
-    frontmatter = {}
-    frontmatter_started = False
-    frontmatter_ended = False
-    frontmatter_pattern = re.compile(r"^\s*([a-z_]+):\s*(.*)\s*$", re.IGNORECASE)
-
-    try:
-        with open(file_path, "r", encoding="utf-8") as file:
-            first_line = file.readline()
-            if first_line.strip() != '"""':
-                # The file doesn't start with triple quotes
-                return {}
-
-            frontmatter_started = True
-
-            for line in file:
-                if '"""' in line:
-                    if frontmatter_started:
-                        frontmatter_ended = True
-                        break
-
-                if frontmatter_started and not frontmatter_ended:
-                    match = frontmatter_pattern.match(line)
-                    if match:
-                        key, value = match.groups()
-                        frontmatter[key.strip()] = value.strip()
-
-    except FileNotFoundError:
-        print(f"Error: The file {file_path} does not exist.")
-        return {}
-    except Exception as e:
-        print(f"An error occurred: {e}")
-        return {}
-
-    return frontmatter
-
-
-def load_toolkit_module_by_id(toolkit_id):
-    toolkit_path = os.path.join(TOOLS_DIR, f"{toolkit_id}.py")
-    spec = util.spec_from_file_location(toolkit_id, toolkit_path)
-    module = util.module_from_spec(spec)
-    frontmatter = extract_frontmatter(toolkit_path)
-
-    try:
-        spec.loader.exec_module(module)
-        print(f"Loaded module: {module.__name__}")
-        if hasattr(module, "Tools"):
-            return module.Tools(), frontmatter
-        else:
-            raise Exception("No Tools class found")
-    except Exception as e:
-        print(f"Error loading module: {toolkit_id}")
-        # Move the file to the error folder
-        os.rename(toolkit_path, f"{toolkit_path}.error")
-        raise e
-
-
-def load_function_module_by_id(function_id):
-    function_path = os.path.join(FUNCTIONS_DIR, f"{function_id}.py")
-
-    spec = util.spec_from_file_location(function_id, function_path)
-    module = util.module_from_spec(spec)
-    frontmatter = extract_frontmatter(function_path)
-
-    try:
-        spec.loader.exec_module(module)
-        print(f"Loaded module: {module.__name__}")
-        if hasattr(module, "Pipe"):
-            return module.Pipe(), "pipe", frontmatter
-        elif hasattr(module, "Filter"):
-            return module.Filter(), "filter", frontmatter
-        elif hasattr(module, "Action"):
-            return module.Action(), "action", frontmatter
-        else:
-            raise Exception("No Function class found")
-    except Exception as e:
-        print(f"Error loading module: {function_id}")
-        # Move the file to the error folder
-        os.rename(function_path, f"{function_path}.error")
-        raise e

backend/data/config.json

@@ -1,36 +0,0 @@
-{
-	"version": 0,
-	"ui": {
-		"default_locale": "",
-		"prompt_suggestions": [
-			{
-				"title": ["Help me study", "vocabulary for a college entrance exam"],
-				"content": "Help me study vocabulary: write a sentence for me to fill in the blank, and I'll try to pick the correct option."
-			},
-			{
-				"title": ["Give me ideas", "for what to do with my kids' art"],
-				"content": "What are 5 creative things I could do with my kids' art? I don't want to throw them away, but it's also so much clutter."
-			},
-			{
-				"title": ["Tell me a fun fact", "about the Roman Empire"],
-				"content": "Tell me a random fun fact about the Roman Empire"
-			},
-			{
-				"title": ["Show me a code snippet", "of a website's sticky header"],
-				"content": "Show me a code snippet of a website's sticky header in CSS and JavaScript."
-			},
-			{
-				"title": ["Explain options trading", "if I'm familiar with buying and selling stocks"],
-				"content": "Explain options trading in simple terms if I'm familiar with buying and selling stocks."
-			},
-			{
-				"title": ["Overcome procrastination", "give me tips"],
-				"content": "Could you start by asking me about instances when I procrastinate the most and then give me some suggestions to overcome it?"
-			},
-			{
-				"title": ["Grammar check", "rewrite it for better readability "],
-				"content": "Check the following sentence for grammar and clarity: \"[sentence]\". Rewrite it for better readability while maintaining its original meaning."
-			}
-		]
-	}
-}

backend/data/litellm/config.yaml

@@ -1,4 +0,0 @@
-general_settings: {}
-litellm_settings: {}
-model_list: []
-router_settings: {}

backend/data/readme.txt

@@ -1 +1 @@
-dir for backend files (db, documents, etc.)
+docker dir for backend files (db, documents, etc.)

backend/dev.sh

@@ -1,2 +1,2 @@
 PORT="${PORT:-8080}"
-uvicorn main:app --port $PORT --host 0.0.0.0 --forwarded-allow-ips '*' --reload
+uvicorn open_webui.main:app --port $PORT --host 0.0.0.0 --forwarded-allow-ips '*' --reload

backend/open_webui/__init__.py

@@ -9,8 +9,6 @@ import uvicorn
 app = typer.Typer()
 
 KEY_FILE = Path.cwd() / ".webui_secret_key"
-if (frontend_build_dir := Path(__file__).parent / "frontend").exists():
-    os.environ["FRONTEND_BUILD_DIR"] = str(frontend_build_dir)
 
 
 @app.command()
@@ -18,6 +16,7 @@ def serve(
     host: str = "0.0.0.0",
     port: int = 8080,
 ):
+    os.environ["FROM_INIT_PY"] = "true"
     if os.getenv("WEBUI_SECRET_KEY") is None:
         typer.echo(
             "Loading WEBUI_SECRET_KEY from file, not provided as an environment variable."
@@ -40,9 +39,22 @@ def serve(
                 "/usr/local/lib/python3.11/site-packages/nvidia/cudnn/lib",
             ]
         )
-    import main  # we need set environment variables before importing main
+        try:
+            import torch
 
-    uvicorn.run(main.app, host=host, port=port, forwarded_allow_ips="*")
+            assert torch.cuda.is_available(), "CUDA not available"
+            typer.echo("CUDA seems to be working")
+        except Exception as e:
+            typer.echo(
+                "Error when testing CUDA but USE_CUDA_DOCKER is true. "
+                "Resetting USE_CUDA_DOCKER to false and removing "
+                f"LD_LIBRARY_PATH modifications: {e}"
+            )
+            os.environ["USE_CUDA_DOCKER"] = "false"
+            os.environ["LD_LIBRARY_PATH"] = ":".join(LD_LIBRARY_PATH)
+    import open_webui.main  # we need set environment variables before importing main
+
+    uvicorn.run(open_webui.main.app, host=host, port=port, forwarded_allow_ips="*")
 
 
 @app.command()
@@ -52,7 +64,11 @@ def dev(
     reload: bool = True,
 ):
     uvicorn.run(
-        "main:app", host=host, port=port, reload=reload, forwarded_allow_ips="*"
+        "open_webui.main:app",
+        host=host,
+        port=port,
+        reload=reload,
+        forwarded_allow_ips="*",
     )
 
 

backend/open_webui/apps/audio/main.py

@@ -1,65 +1,49 @@
-import os
-import logging
-from fastapi import (
-    FastAPI,
-    Request,
-    Depends,
-    HTTPException,
-    status,
-    UploadFile,
-    File,
-    Form,
-)
-from fastapi.responses import StreamingResponse, JSONResponse, FileResponse
-
-from fastapi.middleware.cors import CORSMiddleware
-from pydantic import BaseModel
-
-from typing import List
-import uuid
-import requests
 import hashlib
-from pathlib import Path
 import json
+import logging
+import os
+import uuid
+from functools import lru_cache
+from pathlib import Path
 
-from constants import ERROR_MESSAGES
-from utils.utils import (
-    decode_token,
-    get_current_user,
-    get_verified_user,
-    get_admin_user,
-)
-from utils.misc import calculate_sha256
-
-
-from config import (
-    SRC_LOG_LEVELS,
-    CACHE_DIR,
-    UPLOAD_DIR,
-    WHISPER_MODEL,
-    WHISPER_MODEL_DIR,
-    WHISPER_MODEL_AUTO_UPDATE,
-    DEVICE_TYPE,
-    AUDIO_STT_OPENAI_API_BASE_URL,
-    AUDIO_STT_OPENAI_API_KEY,
-    AUDIO_TTS_OPENAI_API_BASE_URL,
-    AUDIO_TTS_OPENAI_API_KEY,
-    AUDIO_TTS_API_KEY,
+import requests
+from open_webui.config import (
     AUDIO_STT_ENGINE,
     AUDIO_STT_MODEL,
+    AUDIO_STT_OPENAI_API_BASE_URL,
+    AUDIO_STT_OPENAI_API_KEY,
+    AUDIO_TTS_API_KEY,
     AUDIO_TTS_ENGINE,
     AUDIO_TTS_MODEL,
+    AUDIO_TTS_OPENAI_API_BASE_URL,
+    AUDIO_TTS_OPENAI_API_KEY,
+    AUDIO_TTS_SPLIT_ON,
     AUDIO_TTS_VOICE,
+    AUDIO_TTS_AZURE_SPEECH_REGION,
+    AUDIO_TTS_AZURE_SPEECH_OUTPUT_FORMAT,
+    CACHE_DIR,
+    CORS_ALLOW_ORIGIN,
+    WHISPER_MODEL,
+    WHISPER_MODEL_AUTO_UPDATE,
+    WHISPER_MODEL_DIR,
     AppConfig,
 )
 
+from open_webui.constants import ERROR_MESSAGES
+from open_webui.env import SRC_LOG_LEVELS, DEVICE_TYPE
+from fastapi import Depends, FastAPI, File, HTTPException, Request, UploadFile, status
+from fastapi.middleware.cors import CORSMiddleware
+from fastapi.responses import FileResponse
+from pydantic import BaseModel
+from open_webui.utils.utils import get_admin_user, get_current_user, get_verified_user
+
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["AUDIO"])
 
 app = FastAPI()
 app.add_middleware(
     CORSMiddleware,
-    allow_origins=["*"],
+    allow_origins=CORS_ALLOW_ORIGIN,
     allow_credentials=True,
     allow_methods=["*"],
     allow_headers=["*"],
@@ -78,6 +62,10 @@ app.state.config.TTS_ENGINE = AUDIO_TTS_ENGINE
 app.state.config.TTS_MODEL = AUDIO_TTS_MODEL
 app.state.config.TTS_VOICE = AUDIO_TTS_VOICE
 app.state.config.TTS_API_KEY = AUDIO_TTS_API_KEY
+app.state.config.TTS_SPLIT_ON = AUDIO_TTS_SPLIT_ON
+
+app.state.config.TTS_AZURE_SPEECH_REGION = AUDIO_TTS_AZURE_SPEECH_REGION
+app.state.config.TTS_AZURE_SPEECH_OUTPUT_FORMAT = AUDIO_TTS_AZURE_SPEECH_OUTPUT_FORMAT
 
 # setting device type for whisper model
 whisper_device_type = DEVICE_TYPE if DEVICE_TYPE and DEVICE_TYPE == "cuda" else "cpu"
@@ -94,6 +82,9 @@ class TTSConfigForm(BaseModel):
     ENGINE: str
     MODEL: str
     VOICE: str
+    SPLIT_ON: str
+    AZURE_SPEECH_REGION: str
+    AZURE_SPEECH_OUTPUT_FORMAT: str
 
 
 class STTConfigForm(BaseModel):
@@ -145,6 +136,9 @@ async def get_audio_config(user=Depends(get_admin_user)):
             "ENGINE": app.state.config.TTS_ENGINE,
             "MODEL": app.state.config.TTS_MODEL,
             "VOICE": app.state.config.TTS_VOICE,
+            "SPLIT_ON": app.state.config.TTS_SPLIT_ON,
+            "AZURE_SPEECH_REGION": app.state.config.TTS_AZURE_SPEECH_REGION,
+            "AZURE_SPEECH_OUTPUT_FORMAT": app.state.config.TTS_AZURE_SPEECH_OUTPUT_FORMAT,
         },
         "stt": {
             "OPENAI_API_BASE_URL": app.state.config.STT_OPENAI_API_BASE_URL,
@@ -165,6 +159,11 @@ async def update_audio_config(
     app.state.config.TTS_ENGINE = form_data.tts.ENGINE
     app.state.config.TTS_MODEL = form_data.tts.MODEL
     app.state.config.TTS_VOICE = form_data.tts.VOICE
+    app.state.config.TTS_SPLIT_ON = form_data.tts.SPLIT_ON
+    app.state.config.TTS_AZURE_SPEECH_REGION = form_data.tts.AZURE_SPEECH_REGION
+    app.state.config.TTS_AZURE_SPEECH_OUTPUT_FORMAT = (
+        form_data.tts.AZURE_SPEECH_OUTPUT_FORMAT
+    )
 
     app.state.config.STT_OPENAI_API_BASE_URL = form_data.stt.OPENAI_API_BASE_URL
     app.state.config.STT_OPENAI_API_KEY = form_data.stt.OPENAI_API_KEY
@@ -179,6 +178,9 @@ async def update_audio_config(
             "ENGINE": app.state.config.TTS_ENGINE,
             "MODEL": app.state.config.TTS_MODEL,
             "VOICE": app.state.config.TTS_VOICE,
+            "SPLIT_ON": app.state.config.TTS_SPLIT_ON,
+            "AZURE_SPEECH_REGION": app.state.config.TTS_AZURE_SPEECH_REGION,
+            "AZURE_SPEECH_OUTPUT_FORMAT": app.state.config.TTS_AZURE_SPEECH_OUTPUT_FORMAT,
         },
         "stt": {
             "OPENAI_API_BASE_URL": app.state.config.STT_OPENAI_API_BASE_URL,
@@ -211,7 +213,7 @@ async def speech(request: Request, user=Depends(get_verified_user)):
             body = json.loads(body)
             body["model"] = app.state.config.TTS_MODEL
             body = json.dumps(body).encode("utf-8")
-        except Exception as e:
+        except Exception:
             pass
 
         r = None
@@ -244,7 +246,7 @@ async def speech(request: Request, user=Depends(get_verified_user)):
                     res = r.json()
                     if "error" in res:
                         error_detail = f"External: {res['error']['message']}"
-                except:
+                except Exception:
                     error_detail = f"External: {e}"
 
             raise HTTPException(
@@ -261,6 +263,13 @@ async def speech(request: Request, user=Depends(get_verified_user)):
             raise HTTPException(status_code=400, detail="Invalid JSON payload")
 
         voice_id = payload.get("voice", "")
+
+        if voice_id not in get_available_voices():
+            raise HTTPException(
+                status_code=400,
+                detail="Invalid voice id",
+            )
+
         url = f"https://api.elevenlabs.io/v1/text-to-speech/{voice_id}"
 
         headers = {
@@ -299,7 +308,7 @@ async def speech(request: Request, user=Depends(get_verified_user)):
                     res = r.json()
                     if "error" in res:
                         error_detail = f"External: {res['error']['message']}"
-                except:
+                except Exception:
                     error_detail = f"External: {e}"
 
             raise HTTPException(
@@ -307,6 +316,42 @@ async def speech(request: Request, user=Depends(get_verified_user)):
                 detail=error_detail,
             )
 
+    elif app.state.config.TTS_ENGINE == "azure":
+        payload = None
+        try:
+            payload = json.loads(body.decode("utf-8"))
+        except Exception as e:
+            log.exception(e)
+            raise HTTPException(status_code=400, detail="Invalid JSON payload")
+
+        region = app.state.config.TTS_AZURE_SPEECH_REGION
+        language = app.state.config.TTS_VOICE
+        locale = "-".join(app.state.config.TTS_VOICE.split("-")[:1])
+        output_format = app.state.config.TTS_AZURE_SPEECH_OUTPUT_FORMAT
+        url = f"https://{region}.tts.speech.microsoft.com/cognitiveservices/v1"
+
+        headers = {
+            "Ocp-Apim-Subscription-Key": app.state.config.TTS_API_KEY,
+            "Content-Type": "application/ssml+xml",
+            "X-Microsoft-OutputFormat": output_format,
+        }
+
+        data = f"""<speak version="1.0" xmlns="http://www.w3.org/2001/10/synthesis" xml:lang="{locale}">
+                <voice name="{language}">{payload["input"]}</voice>
+            </speak>"""
+
+        response = requests.post(url, headers=headers, data=data)
+
+        if response.status_code == 200:
+            with open(file_path, "wb") as f:
+                f.write(response.content)
+            return FileResponse(file_path)
+        else:
+            log.error(f"Error synthesizing speech - {response.reason}")
+            raise HTTPException(
+                status_code=500, detail=f"Error synthesizing speech - {response.reason}"
+            )
+
 
 @app.post("/transcriptions")
 def transcribe(
@@ -315,7 +360,7 @@ def transcribe(
 ):
     log.info(f"file.content_type: {file.content_type}")
 
-    if file.content_type not in ["audio/mpeg", "audio/wav"]:
+    if file.content_type not in ["audio/mpeg", "audio/wav", "audio/ogg"]:
         raise HTTPException(
             status_code=status.HTTP_400_BAD_REQUEST,
             detail=ERROR_MESSAGES.FILE_NOT_SUPPORTED,
@@ -353,7 +398,7 @@ def transcribe(
 
             try:
                 model = WhisperModel(**whisper_kwargs)
-            except:
+            except Exception:
                 log.warning(
                     "WhisperModel initialization failed, attempting download with local_files_only=False"
                 )
@@ -421,7 +466,7 @@ def transcribe(
                         res = r.json()
                         if "error" in res:
                             error_detail = f"External: {res['error']['message']}"
-                    except:
+                    except Exception:
                         error_detail = f"External: {e}"
 
                 raise HTTPException(
@@ -438,7 +483,7 @@ def transcribe(
         )
 
 
-def get_available_models() -> List[dict]:
+def get_available_models() -> list[dict]:
     if app.state.config.TTS_ENGINE == "openai":
         return [{"id": "tts-1"}, {"id": "tts-1-hd"}]
     elif app.state.config.TTS_ENGINE == "elevenlabs":
@@ -449,7 +494,7 @@ def get_available_models() -> List[dict]:
 
         try:
             response = requests.get(
-                "https://api.elevenlabs.io/v1/models", headers=headers
+                "https://api.elevenlabs.io/v1/models", headers=headers, timeout=5
             )
             response.raise_for_status()
             models = response.json()
@@ -466,39 +511,73 @@ async def get_models(user=Depends(get_verified_user)):
     return {"models": get_available_models()}
 
 
-def get_available_voices() -> List[dict]:
+def get_available_voices() -> dict:
+    """Returns {voice_id: voice_name} dict"""
+    ret = {}
     if app.state.config.TTS_ENGINE == "openai":
-        return [
-            {"name": "alloy", "id": "alloy"},
-            {"name": "echo", "id": "echo"},
-            {"name": "fable", "id": "fable"},
-            {"name": "onyx", "id": "onyx"},
-            {"name": "nova", "id": "nova"},
-            {"name": "shimmer", "id": "shimmer"},
-        ]
-    elif app.state.config.TTS_ENGINE == "elevenlabs":
-        headers = {
-            "xi-api-key": app.state.config.TTS_API_KEY,
-            "Content-Type": "application/json",
+        ret = {
+            "alloy": "alloy",
+            "echo": "echo",
+            "fable": "fable",
+            "onyx": "onyx",
+            "nova": "nova",
+            "shimmer": "shimmer",
         }
-
+    elif app.state.config.TTS_ENGINE == "elevenlabs":
         try:
-            response = requests.get(
-                "https://api.elevenlabs.io/v1/voices", headers=headers
-            )
-            response.raise_for_status()
-            voices_data = response.json()
+            ret = get_elevenlabs_voices()
+        except Exception:
+            # Avoided @lru_cache with exception
+            pass
+    elif app.state.config.TTS_ENGINE == "azure":
+        try:
+            region = app.state.config.TTS_AZURE_SPEECH_REGION
+            url = f"https://{region}.tts.speech.microsoft.com/cognitiveservices/voices/list"
+            headers = {"Ocp-Apim-Subscription-Key": app.state.config.TTS_API_KEY}
 
-            voices = []
-            for voice in voices_data.get("voices", []):
-                voices.append({"name": voice["name"], "id": voice["voice_id"]})
-            return voices
+            response = requests.get(url, headers=headers)
+            response.raise_for_status()
+            voices = response.json()
+            for voice in voices:
+                ret[voice["ShortName"]] = (
+                    f"{voice['DisplayName']} ({voice['ShortName']})"
+                )
         except requests.RequestException as e:
             log.error(f"Error fetching voices: {str(e)}")
 
-    return []
+    return ret
+
+
+@lru_cache
+def get_elevenlabs_voices() -> dict:
+    """
+    Note, set the following in your .env file to use Elevenlabs:
+    AUDIO_TTS_ENGINE=elevenlabs
+    AUDIO_TTS_API_KEY=sk_...  # Your Elevenlabs API key
+    AUDIO_TTS_VOICE=EXAVITQu4vr4xnSDxMaL  # From https://api.elevenlabs.io/v1/voices
+    AUDIO_TTS_MODEL=eleven_multilingual_v2
+    """
+    headers = {
+        "xi-api-key": app.state.config.TTS_API_KEY,
+        "Content-Type": "application/json",
+    }
+    try:
+        # TODO: Add retries
+        response = requests.get("https://api.elevenlabs.io/v1/voices", headers=headers)
+        response.raise_for_status()
+        voices_data = response.json()
+
+        voices = {}
+        for voice in voices_data.get("voices", []):
+            voices[voice["voice_id"]] = voice["name"]
+    except requests.RequestException as e:
+        # Avoid @lru_cache with exception
+        log.error(f"Error fetching voices: {str(e)}")
+        raise RuntimeError(f"Error fetching voices: {str(e)}")
+
+    return voices
 
 
 @app.get("/voices")
 async def get_voices(user=Depends(get_verified_user)):
-    return {"voices": get_available_voices()}
+    return {"voices": [{"id": k, "name": v} for k, v in get_available_voices().items()]}

backend/open_webui/apps/images/main.py

@@ -1,57 +1,45 @@
-import re
-import requests
-import base64
-from fastapi import (
-    FastAPI,
-    Request,
-    Depends,
-    HTTPException,
-    status,
-    UploadFile,
-    File,
-    Form,
-)
-from fastapi.middleware.cors import CORSMiddleware
-
-from constants import ERROR_MESSAGES
-from utils.utils import (
-    get_verified_user,
-    get_admin_user,
-)
-
-from apps.images.utils.comfyui import ImageGenerationPayload, comfyui_generate_image
-from utils.misc import calculate_sha256
-from typing import Optional
-from pydantic import BaseModel
-from pathlib import Path
-import mimetypes
-import uuid
+import asyncio
 import base64
 import json
 import logging
+import mimetypes
+import re
+import uuid
+from pathlib import Path
+from typing import Optional
 
-from config import (
-    SRC_LOG_LEVELS,
-    CACHE_DIR,
-    IMAGE_GENERATION_ENGINE,
-    ENABLE_IMAGE_GENERATION,
-    AUTOMATIC1111_BASE_URL,
+import requests
+from open_webui.apps.images.utils.comfyui import (
+    ComfyUIGenerateImageForm,
+    ComfyUIWorkflow,
+    comfyui_generate_image,
+)
+from open_webui.config import (
     AUTOMATIC1111_API_AUTH,
+    AUTOMATIC1111_BASE_URL,
+    AUTOMATIC1111_CFG_SCALE,
+    AUTOMATIC1111_SAMPLER,
+    AUTOMATIC1111_SCHEDULER,
+    CACHE_DIR,
     COMFYUI_BASE_URL,
-    COMFYUI_CFG_SCALE,
-    COMFYUI_SAMPLER,
-    COMFYUI_SCHEDULER,
-    COMFYUI_SD3,
-    COMFYUI_FLUX,
-    COMFYUI_FLUX_WEIGHT_DTYPE,
-    COMFYUI_FLUX_FP8_CLIP,
-    IMAGES_OPENAI_API_BASE_URL,
-    IMAGES_OPENAI_API_KEY,
+    COMFYUI_WORKFLOW,
+    COMFYUI_WORKFLOW_NODES,
+    CORS_ALLOW_ORIGIN,
+    ENABLE_IMAGE_GENERATION,
+    IMAGE_GENERATION_ENGINE,
     IMAGE_GENERATION_MODEL,
     IMAGE_SIZE,
     IMAGE_STEPS,
+    IMAGES_OPENAI_API_BASE_URL,
+    IMAGES_OPENAI_API_KEY,
     AppConfig,
 )
+from open_webui.constants import ERROR_MESSAGES
+from open_webui.env import SRC_LOG_LEVELS
+from fastapi import Depends, FastAPI, HTTPException, Request
+from fastapi.middleware.cors import CORSMiddleware
+from pydantic import BaseModel
+from open_webui.utils.utils import get_admin_user, get_verified_user
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["IMAGES"])
@@ -62,7 +50,7 @@ IMAGE_CACHE_DIR.mkdir(parents=True, exist_ok=True)
 app = FastAPI()
 app.add_middleware(
     CORSMiddleware,
-    allow_origins=["*"],
+    allow_origins=CORS_ALLOW_ORIGIN,
     allow_credentials=True,
     allow_methods=["*"],
     allow_headers=["*"],
@@ -80,21 +68,127 @@ app.state.config.MODEL = IMAGE_GENERATION_MODEL
 
 app.state.config.AUTOMATIC1111_BASE_URL = AUTOMATIC1111_BASE_URL
 app.state.config.AUTOMATIC1111_API_AUTH = AUTOMATIC1111_API_AUTH
+app.state.config.AUTOMATIC1111_CFG_SCALE = AUTOMATIC1111_CFG_SCALE
+app.state.config.AUTOMATIC1111_SAMPLER = AUTOMATIC1111_SAMPLER
+app.state.config.AUTOMATIC1111_SCHEDULER = AUTOMATIC1111_SCHEDULER
 app.state.config.COMFYUI_BASE_URL = COMFYUI_BASE_URL
+app.state.config.COMFYUI_WORKFLOW = COMFYUI_WORKFLOW
+app.state.config.COMFYUI_WORKFLOW_NODES = COMFYUI_WORKFLOW_NODES
 
 app.state.config.IMAGE_SIZE = IMAGE_SIZE
 app.state.config.IMAGE_STEPS = IMAGE_STEPS
-app.state.config.COMFYUI_CFG_SCALE = COMFYUI_CFG_SCALE
-app.state.config.COMFYUI_SAMPLER = COMFYUI_SAMPLER
-app.state.config.COMFYUI_SCHEDULER = COMFYUI_SCHEDULER
-app.state.config.COMFYUI_SD3 = COMFYUI_SD3
-app.state.config.COMFYUI_FLUX = COMFYUI_FLUX
-app.state.config.COMFYUI_FLUX_WEIGHT_DTYPE = COMFYUI_FLUX_WEIGHT_DTYPE
-app.state.config.COMFYUI_FLUX_FP8_CLIP = COMFYUI_FLUX_FP8_CLIP
+
+
+@app.get("/config")
+async def get_config(request: Request, user=Depends(get_admin_user)):
+    return {
+        "enabled": app.state.config.ENABLED,
+        "engine": app.state.config.ENGINE,
+        "openai": {
+            "OPENAI_API_BASE_URL": app.state.config.OPENAI_API_BASE_URL,
+            "OPENAI_API_KEY": app.state.config.OPENAI_API_KEY,
+        },
+        "automatic1111": {
+            "AUTOMATIC1111_BASE_URL": app.state.config.AUTOMATIC1111_BASE_URL,
+            "AUTOMATIC1111_API_AUTH": app.state.config.AUTOMATIC1111_API_AUTH,
+            "AUTOMATIC1111_CFG_SCALE": app.state.config.AUTOMATIC1111_CFG_SCALE,
+            "AUTOMATIC1111_SAMPLER": app.state.config.AUTOMATIC1111_SAMPLER,
+            "AUTOMATIC1111_SCHEDULER": app.state.config.AUTOMATIC1111_SCHEDULER,
+        },
+        "comfyui": {
+            "COMFYUI_BASE_URL": app.state.config.COMFYUI_BASE_URL,
+            "COMFYUI_WORKFLOW": app.state.config.COMFYUI_WORKFLOW,
+            "COMFYUI_WORKFLOW_NODES": app.state.config.COMFYUI_WORKFLOW_NODES,
+        },
+    }
+
+
+class OpenAIConfigForm(BaseModel):
+    OPENAI_API_BASE_URL: str
+    OPENAI_API_KEY: str
+
+
+class Automatic1111ConfigForm(BaseModel):
+    AUTOMATIC1111_BASE_URL: str
+    AUTOMATIC1111_API_AUTH: str
+    AUTOMATIC1111_CFG_SCALE: Optional[str]
+    AUTOMATIC1111_SAMPLER: Optional[str]
+    AUTOMATIC1111_SCHEDULER: Optional[str]
+
+
+class ComfyUIConfigForm(BaseModel):
+    COMFYUI_BASE_URL: str
+    COMFYUI_WORKFLOW: str
+    COMFYUI_WORKFLOW_NODES: list[dict]
+
+
+class ConfigForm(BaseModel):
+    enabled: bool
+    engine: str
+    openai: OpenAIConfigForm
+    automatic1111: Automatic1111ConfigForm
+    comfyui: ComfyUIConfigForm
+
+
+@app.post("/config/update")
+async def update_config(form_data: ConfigForm, user=Depends(get_admin_user)):
+    app.state.config.ENGINE = form_data.engine
+    app.state.config.ENABLED = form_data.enabled
+
+    app.state.config.OPENAI_API_BASE_URL = form_data.openai.OPENAI_API_BASE_URL
+    app.state.config.OPENAI_API_KEY = form_data.openai.OPENAI_API_KEY
+
+    app.state.config.AUTOMATIC1111_BASE_URL = (
+        form_data.automatic1111.AUTOMATIC1111_BASE_URL
+    )
+    app.state.config.AUTOMATIC1111_API_AUTH = (
+        form_data.automatic1111.AUTOMATIC1111_API_AUTH
+    )
+
+    app.state.config.AUTOMATIC1111_CFG_SCALE = (
+        float(form_data.automatic1111.AUTOMATIC1111_CFG_SCALE)
+        if form_data.automatic1111.AUTOMATIC1111_CFG_SCALE
+        else None
+    )
+    app.state.config.AUTOMATIC1111_SAMPLER = (
+        form_data.automatic1111.AUTOMATIC1111_SAMPLER
+        if form_data.automatic1111.AUTOMATIC1111_SAMPLER
+        else None
+    )
+    app.state.config.AUTOMATIC1111_SCHEDULER = (
+        form_data.automatic1111.AUTOMATIC1111_SCHEDULER
+        if form_data.automatic1111.AUTOMATIC1111_SCHEDULER
+        else None
+    )
+
+    app.state.config.COMFYUI_BASE_URL = form_data.comfyui.COMFYUI_BASE_URL.strip("/")
+    app.state.config.COMFYUI_WORKFLOW = form_data.comfyui.COMFYUI_WORKFLOW
+    app.state.config.COMFYUI_WORKFLOW_NODES = form_data.comfyui.COMFYUI_WORKFLOW_NODES
+
+    return {
+        "enabled": app.state.config.ENABLED,
+        "engine": app.state.config.ENGINE,
+        "openai": {
+            "OPENAI_API_BASE_URL": app.state.config.OPENAI_API_BASE_URL,
+            "OPENAI_API_KEY": app.state.config.OPENAI_API_KEY,
+        },
+        "automatic1111": {
+            "AUTOMATIC1111_BASE_URL": app.state.config.AUTOMATIC1111_BASE_URL,
+            "AUTOMATIC1111_API_AUTH": app.state.config.AUTOMATIC1111_API_AUTH,
+            "AUTOMATIC1111_CFG_SCALE": app.state.config.AUTOMATIC1111_CFG_SCALE,
+            "AUTOMATIC1111_SAMPLER": app.state.config.AUTOMATIC1111_SAMPLER,
+            "AUTOMATIC1111_SCHEDULER": app.state.config.AUTOMATIC1111_SCHEDULER,
+        },
+        "comfyui": {
+            "COMFYUI_BASE_URL": app.state.config.COMFYUI_BASE_URL,
+            "COMFYUI_WORKFLOW": app.state.config.COMFYUI_WORKFLOW,
+            "COMFYUI_WORKFLOW_NODES": app.state.config.COMFYUI_WORKFLOW_NODES,
+        },
+    }
 
 
 def get_automatic1111_api_auth():
-    if app.state.config.AUTOMATIC1111_API_AUTH == None:
+    if app.state.config.AUTOMATIC1111_API_AUTH is None:
         return ""
     else:
         auth1111_byte_string = app.state.config.AUTOMATIC1111_API_AUTH.encode("utf-8")
@@ -103,164 +197,112 @@ def get_automatic1111_api_auth():
         return f"Basic {auth1111_base64_encoded_string}"
 
 
-@app.get("/config")
-async def get_config(request: Request, user=Depends(get_admin_user)):
-    return {
-        "engine": app.state.config.ENGINE,
-        "enabled": app.state.config.ENABLED,
-    }
-
-
-class ConfigUpdateForm(BaseModel):
-    engine: str
-    enabled: bool
-
-
-@app.post("/config/update")
-async def update_config(form_data: ConfigUpdateForm, user=Depends(get_admin_user)):
-    app.state.config.ENGINE = form_data.engine
-    app.state.config.ENABLED = form_data.enabled
-    return {
-        "engine": app.state.config.ENGINE,
-        "enabled": app.state.config.ENABLED,
-    }
-
-
-class EngineUrlUpdateForm(BaseModel):
-    AUTOMATIC1111_BASE_URL: Optional[str] = None
-    AUTOMATIC1111_API_AUTH: Optional[str] = None
-    COMFYUI_BASE_URL: Optional[str] = None
-
-
-@app.get("/url")
-async def get_engine_url(user=Depends(get_admin_user)):
-    return {
-        "AUTOMATIC1111_BASE_URL": app.state.config.AUTOMATIC1111_BASE_URL,
-        "AUTOMATIC1111_API_AUTH": app.state.config.AUTOMATIC1111_API_AUTH,
-        "COMFYUI_BASE_URL": app.state.config.COMFYUI_BASE_URL,
-    }
-
-
-@app.post("/url/update")
-async def update_engine_url(
-    form_data: EngineUrlUpdateForm, user=Depends(get_admin_user)
-):
-    if form_data.AUTOMATIC1111_BASE_URL == None:
-        app.state.config.AUTOMATIC1111_BASE_URL = AUTOMATIC1111_BASE_URL
-    else:
-        url = form_data.AUTOMATIC1111_BASE_URL.strip("/")
+@app.get("/config/url/verify")
+async def verify_url(user=Depends(get_admin_user)):
+    if app.state.config.ENGINE == "automatic1111":
         try:
-            r = requests.head(url)
-            app.state.config.AUTOMATIC1111_BASE_URL = url
+            r = requests.get(
+                url=f"{app.state.config.AUTOMATIC1111_BASE_URL}/sdapi/v1/options",
+                headers={"authorization": get_automatic1111_api_auth()},
+            )
+            r.raise_for_status()
+            return True
+        except Exception:
+            app.state.config.ENABLED = False
+            raise HTTPException(status_code=400, detail=ERROR_MESSAGES.INVALID_URL)
+    elif app.state.config.ENGINE == "comfyui":
+        try:
+            r = requests.get(url=f"{app.state.config.COMFYUI_BASE_URL}/object_info")
+            r.raise_for_status()
+            return True
+        except Exception:
+            app.state.config.ENABLED = False
+            raise HTTPException(status_code=400, detail=ERROR_MESSAGES.INVALID_URL)
+    else:
+        return True
+
+
+def set_image_model(model: str):
+    log.info(f"Setting image model to {model}")
+    app.state.config.MODEL = model
+    if app.state.config.ENGINE in ["", "automatic1111"]:
+        api_auth = get_automatic1111_api_auth()
+        r = requests.get(
+            url=f"{app.state.config.AUTOMATIC1111_BASE_URL}/sdapi/v1/options",
+            headers={"authorization": api_auth},
+        )
+        options = r.json()
+        if model != options["sd_model_checkpoint"]:
+            options["sd_model_checkpoint"] = model
+            r = requests.post(
+                url=f"{app.state.config.AUTOMATIC1111_BASE_URL}/sdapi/v1/options",
+                json=options,
+                headers={"authorization": api_auth},
+            )
+    return app.state.config.MODEL
+
+
+def get_image_model():
+    if app.state.config.ENGINE == "openai":
+        return app.state.config.MODEL if app.state.config.MODEL else "dall-e-2"
+    elif app.state.config.ENGINE == "comfyui":
+        return app.state.config.MODEL if app.state.config.MODEL else ""
+    elif app.state.config.ENGINE == "automatic1111" or app.state.config.ENGINE == "":
+        try:
+            r = requests.get(
+                url=f"{app.state.config.AUTOMATIC1111_BASE_URL}/sdapi/v1/options",
+                headers={"authorization": get_automatic1111_api_auth()},
+            )
+            options = r.json()
+            return options["sd_model_checkpoint"]
         except Exception as e:
+            app.state.config.ENABLED = False
             raise HTTPException(status_code=400, detail=ERROR_MESSAGES.DEFAULT(e))
 
-    if form_data.COMFYUI_BASE_URL == None:
-        app.state.config.COMFYUI_BASE_URL = COMFYUI_BASE_URL
-    else:
-        url = form_data.COMFYUI_BASE_URL.strip("/")
 
-        try:
-            r = requests.head(url)
-            app.state.config.COMFYUI_BASE_URL = url
-        except Exception as e:
-            raise HTTPException(status_code=400, detail=ERROR_MESSAGES.DEFAULT(e))
+class ImageConfigForm(BaseModel):
+    MODEL: str
+    IMAGE_SIZE: str
+    IMAGE_STEPS: int
 
-    if form_data.AUTOMATIC1111_API_AUTH == None:
-        app.state.config.AUTOMATIC1111_API_AUTH = AUTOMATIC1111_API_AUTH
-    else:
-        app.state.config.AUTOMATIC1111_API_AUTH = form_data.AUTOMATIC1111_API_AUTH
 
+@app.get("/image/config")
+async def get_image_config(user=Depends(get_admin_user)):
     return {
-        "AUTOMATIC1111_BASE_URL": app.state.config.AUTOMATIC1111_BASE_URL,
-        "AUTOMATIC1111_API_AUTH": app.state.config.AUTOMATIC1111_API_AUTH,
-        "COMFYUI_BASE_URL": app.state.config.COMFYUI_BASE_URL,
-        "status": True,
+        "MODEL": app.state.config.MODEL,
+        "IMAGE_SIZE": app.state.config.IMAGE_SIZE,
+        "IMAGE_STEPS": app.state.config.IMAGE_STEPS,
     }
 
 
-class OpenAIConfigUpdateForm(BaseModel):
-    url: str
-    key: str
+@app.post("/image/config/update")
+async def update_image_config(form_data: ImageConfigForm, user=Depends(get_admin_user)):
 
+    set_image_model(form_data.MODEL)
 
-@app.get("/openai/config")
-async def get_openai_config(user=Depends(get_admin_user)):
-    return {
-        "OPENAI_API_BASE_URL": app.state.config.OPENAI_API_BASE_URL,
-        "OPENAI_API_KEY": app.state.config.OPENAI_API_KEY,
-    }
-
-
-@app.post("/openai/config/update")
-async def update_openai_config(
-    form_data: OpenAIConfigUpdateForm, user=Depends(get_admin_user)
-):
-    if form_data.key == "":
-        raise HTTPException(status_code=400, detail=ERROR_MESSAGES.API_KEY_NOT_FOUND)
-
-    app.state.config.OPENAI_API_BASE_URL = form_data.url
-    app.state.config.OPENAI_API_KEY = form_data.key
-
-    return {
-        "status": True,
-        "OPENAI_API_BASE_URL": app.state.config.OPENAI_API_BASE_URL,
-        "OPENAI_API_KEY": app.state.config.OPENAI_API_KEY,
-    }
-
-
-class ImageSizeUpdateForm(BaseModel):
-    size: str
-
-
-@app.get("/size")
-async def get_image_size(user=Depends(get_admin_user)):
-    return {"IMAGE_SIZE": app.state.config.IMAGE_SIZE}
-
-
-@app.post("/size/update")
-async def update_image_size(
-    form_data: ImageSizeUpdateForm, user=Depends(get_admin_user)
-):
-    pattern = r"^\d+x\d+$"  # Regular expression pattern
-    if re.match(pattern, form_data.size):
-        app.state.config.IMAGE_SIZE = form_data.size
-        return {
-            "IMAGE_SIZE": app.state.config.IMAGE_SIZE,
-            "status": True,
-        }
+    pattern = r"^\d+x\d+$"
+    if re.match(pattern, form_data.IMAGE_SIZE):
+        app.state.config.IMAGE_SIZE = form_data.IMAGE_SIZE
     else:
         raise HTTPException(
             status_code=400,
             detail=ERROR_MESSAGES.INCORRECT_FORMAT("  (e.g., 512x512)."),
         )
 
-
-class ImageStepsUpdateForm(BaseModel):
-    steps: int
-
-
-@app.get("/steps")
-async def get_image_size(user=Depends(get_admin_user)):
-    return {"IMAGE_STEPS": app.state.config.IMAGE_STEPS}
-
-
-@app.post("/steps/update")
-async def update_image_size(
-    form_data: ImageStepsUpdateForm, user=Depends(get_admin_user)
-):
-    if form_data.steps >= 0:
-        app.state.config.IMAGE_STEPS = form_data.steps
-        return {
-            "IMAGE_STEPS": app.state.config.IMAGE_STEPS,
-            "status": True,
-        }
+    if form_data.IMAGE_STEPS >= 0:
+        app.state.config.IMAGE_STEPS = form_data.IMAGE_STEPS
     else:
         raise HTTPException(
             status_code=400,
             detail=ERROR_MESSAGES.INCORRECT_FORMAT("  (e.g., 50)."),
         )
 
+    return {
+        "MODEL": app.state.config.MODEL,
+        "IMAGE_SIZE": app.state.config.IMAGE_SIZE,
+        "IMAGE_STEPS": app.state.config.IMAGE_STEPS,
+    }
+
 
 @app.get("/models")
 def get_models(user=Depends(get_verified_user)):
@@ -271,18 +313,51 @@ def get_models(user=Depends(get_verified_user)):
                 {"id": "dall-e-3", "name": "DALL·E 3"},
             ]
         elif app.state.config.ENGINE == "comfyui":
-
+            # TODO - get models from comfyui
             r = requests.get(url=f"{app.state.config.COMFYUI_BASE_URL}/object_info")
             info = r.json()
 
-            return list(
-                map(
-                    lambda model: {"id": model, "name": model},
-                    info["CheckpointLoaderSimple"]["input"]["required"]["ckpt_name"][0],
-                )
-            )
+            workflow = json.loads(app.state.config.COMFYUI_WORKFLOW)
+            model_node_id = None
 
-        else:
+            for node in app.state.config.COMFYUI_WORKFLOW_NODES:
+                if node["type"] == "model":
+                    if node["node_ids"]:
+                        model_node_id = node["node_ids"][0]
+                    break
+
+            if model_node_id:
+                model_list_key = None
+
+                print(workflow[model_node_id]["class_type"])
+                for key in info[workflow[model_node_id]["class_type"]]["input"][
+                    "required"
+                ]:
+                    if "_name" in key:
+                        model_list_key = key
+                        break
+
+                if model_list_key:
+                    return list(
+                        map(
+                            lambda model: {"id": model, "name": model},
+                            info[workflow[model_node_id]["class_type"]]["input"][
+                                "required"
+                            ][model_list_key][0],
+                        )
+                    )
+            else:
+                return list(
+                    map(
+                        lambda model: {"id": model, "name": model},
+                        info["CheckpointLoaderSimple"]["input"]["required"][
+                            "ckpt_name"
+                        ][0],
+                    )
+                )
+        elif (
+            app.state.config.ENGINE == "automatic1111" or app.state.config.ENGINE == ""
+        ):
             r = requests.get(
                 url=f"{app.state.config.AUTOMATIC1111_BASE_URL}/sdapi/v1/sd-models",
                 headers={"authorization": get_automatic1111_api_auth()},
@@ -299,69 +374,11 @@ def get_models(user=Depends(get_verified_user)):
         raise HTTPException(status_code=400, detail=ERROR_MESSAGES.DEFAULT(e))
 
 
-@app.get("/models/default")
-async def get_default_model(user=Depends(get_admin_user)):
-    try:
-        if app.state.config.ENGINE == "openai":
-            return {
-                "model": (
-                    app.state.config.MODEL if app.state.config.MODEL else "dall-e-2"
-                )
-            }
-        elif app.state.config.ENGINE == "comfyui":
-            return {"model": (app.state.config.MODEL if app.state.config.MODEL else "")}
-        else:
-            r = requests.get(
-                url=f"{app.state.config.AUTOMATIC1111_BASE_URL}/sdapi/v1/options",
-                headers={"authorization": get_automatic1111_api_auth()},
-            )
-            options = r.json()
-            return {"model": options["sd_model_checkpoint"]}
-    except Exception as e:
-        app.state.config.ENABLED = False
-        raise HTTPException(status_code=400, detail=ERROR_MESSAGES.DEFAULT(e))
-
-
-class UpdateModelForm(BaseModel):
-    model: str
-
-
-def set_model_handler(model: str):
-    if app.state.config.ENGINE in ["openai", "comfyui"]:
-        app.state.config.MODEL = model
-        return app.state.config.MODEL
-    else:
-        api_auth = get_automatic1111_api_auth()
-        r = requests.get(
-            url=f"{app.state.config.AUTOMATIC1111_BASE_URL}/sdapi/v1/options",
-            headers={"authorization": api_auth},
-        )
-        options = r.json()
-
-        if model != options["sd_model_checkpoint"]:
-            options["sd_model_checkpoint"] = model
-            r = requests.post(
-                url=f"{app.state.config.AUTOMATIC1111_BASE_URL}/sdapi/v1/options",
-                json=options,
-                headers={"authorization": api_auth},
-            )
-
-        return options
-
-
-@app.post("/models/default/update")
-def update_default_model(
-    form_data: UpdateModelForm,
-    user=Depends(get_verified_user),
-):
-    return set_model_handler(form_data.model)
-
-
 class GenerateImageForm(BaseModel):
     model: Optional[str] = None
     prompt: str
-    n: int = 1
     size: Optional[str] = None
+    n: int = 1
     negative_prompt: Optional[str] = None
 
 
@@ -403,7 +420,6 @@ def save_url_image(url):
         r = requests.get(url)
         r.raise_for_status()
         if r.headers["content-type"].split("/")[0] == "image":
-
             mime_type = r.headers["content-type"]
             image_format = mimetypes.guess_extension(mime_type)
 
@@ -418,7 +434,7 @@ def save_url_image(url):
                     image_file.write(chunk)
             return image_filename
         else:
-            log.error(f"Url does not point to an image.")
+            log.error("Url does not point to an image.")
             return None
 
     except Exception as e:
@@ -436,7 +452,6 @@ async def image_generations(
     r = None
     try:
         if app.state.config.ENGINE == "openai":
-
             headers = {}
             headers["Authorization"] = f"Bearer {app.state.config.OPENAI_API_KEY}"
             headers["Content-Type"] = "application/json"
@@ -455,7 +470,9 @@ async def image_generations(
                 "response_format": "b64_json",
             }
 
-            r = requests.post(
+            # Use asyncio.to_thread for the requests.post call
+            r = await asyncio.to_thread(
+                requests.post,
                 url=f"{app.state.config.OPENAI_API_BASE_URL}/images/generations",
                 json=data,
                 headers=headers,
@@ -477,7 +494,6 @@ async def image_generations(
             return images
 
         elif app.state.config.ENGINE == "comfyui":
-
             data = {
                 "prompt": form_data.prompt,
                 "width": width,
@@ -491,32 +507,20 @@ async def image_generations(
             if form_data.negative_prompt is not None:
                 data["negative_prompt"] = form_data.negative_prompt
 
-            if app.state.config.COMFYUI_CFG_SCALE:
-                data["cfg_scale"] = app.state.config.COMFYUI_CFG_SCALE
-
-            if app.state.config.COMFYUI_SAMPLER is not None:
-                data["sampler"] = app.state.config.COMFYUI_SAMPLER
-
-            if app.state.config.COMFYUI_SCHEDULER is not None:
-                data["scheduler"] = app.state.config.COMFYUI_SCHEDULER
-
-            if app.state.config.COMFYUI_SD3 is not None:
-                data["sd3"] = app.state.config.COMFYUI_SD3
-
-            if app.state.config.COMFYUI_FLUX is not None:
-                data["flux"] = app.state.config.COMFYUI_FLUX
-
-            if app.state.config.COMFYUI_FLUX_WEIGHT_DTYPE is not None:
-                data["flux_weight_dtype"] = app.state.config.COMFYUI_FLUX_WEIGHT_DTYPE
-
-            if app.state.config.COMFYUI_FLUX_FP8_CLIP is not None:
-                data["flux_fp8_clip"] = app.state.config.COMFYUI_FLUX_FP8_CLIP
-
-            data = ImageGenerationPayload(**data)
-
-            res = comfyui_generate_image(
+            form_data = ComfyUIGenerateImageForm(
+                **{
+                    "workflow": ComfyUIWorkflow(
+                        **{
+                            "workflow": app.state.config.COMFYUI_WORKFLOW,
+                            "nodes": app.state.config.COMFYUI_WORKFLOW_NODES,
+                        }
+                    ),
+                    **data,
+                }
+            )
+            res = await comfyui_generate_image(
                 app.state.config.MODEL,
-                data,
+                form_data,
                 user.id,
                 app.state.config.COMFYUI_BASE_URL,
             )
@@ -530,13 +534,15 @@ async def image_generations(
                 file_body_path = IMAGE_CACHE_DIR.joinpath(f"{image_filename}.json")
 
                 with open(file_body_path, "w") as f:
-                    json.dump(data.model_dump(exclude_none=True), f)
+                    json.dump(form_data.model_dump(exclude_none=True), f)
 
             log.debug(f"images: {images}")
             return images
-        else:
+        elif (
+            app.state.config.ENGINE == "automatic1111" or app.state.config.ENGINE == ""
+        ):
             if form_data.model:
-                set_model_handler(form_data.model)
+                set_image_model(form_data.model)
 
             data = {
                 "prompt": form_data.prompt,
@@ -551,14 +557,24 @@ async def image_generations(
             if form_data.negative_prompt is not None:
                 data["negative_prompt"] = form_data.negative_prompt
 
-            r = requests.post(
+            if app.state.config.AUTOMATIC1111_CFG_SCALE:
+                data["cfg_scale"] = app.state.config.AUTOMATIC1111_CFG_SCALE
+
+            if app.state.config.AUTOMATIC1111_SAMPLER:
+                data["sampler_name"] = app.state.config.AUTOMATIC1111_SAMPLER
+
+            if app.state.config.AUTOMATIC1111_SCHEDULER:
+                data["scheduler"] = app.state.config.AUTOMATIC1111_SCHEDULER
+
+            # Use asyncio.to_thread for the requests.post call
+            r = await asyncio.to_thread(
+                requests.post,
                 url=f"{app.state.config.AUTOMATIC1111_BASE_URL}/sdapi/v1/txt2img",
                 json=data,
                 headers={"authorization": get_automatic1111_api_auth()},
             )
 
             res = r.json()
-
             log.debug(f"res: {res}")
 
             images = []
@@ -572,10 +588,8 @@ async def image_generations(
                     json.dump({**data, "info": res["info"]}, f)
 
             return images
-
     except Exception as e:
         error = e
-
         if r != None:
             data = r.json()
             if "error" in data:

backend/open_webui/apps/images/utils/comfyui.py

@@ -0,0 +1,174 @@
+import asyncio
+import json
+import logging
+import random
+import urllib.parse
+import urllib.request
+from typing import Optional
+
+import websocket  # NOTE: websocket-client (https://github.com/websocket-client/websocket-client)
+from open_webui.env import SRC_LOG_LEVELS
+from pydantic import BaseModel
+
+log = logging.getLogger(__name__)
+log.setLevel(SRC_LOG_LEVELS["COMFYUI"])
+
+default_headers = {"User-Agent": "Mozilla/5.0"}
+
+
+def queue_prompt(prompt, client_id, base_url):
+    log.info("queue_prompt")
+    p = {"prompt": prompt, "client_id": client_id}
+    data = json.dumps(p).encode("utf-8")
+    log.debug(f"queue_prompt data: {data}")
+    try:
+        req = urllib.request.Request(
+            f"{base_url}/prompt", data=data, headers=default_headers
+        )
+        response = urllib.request.urlopen(req).read()
+        return json.loads(response)
+    except Exception as e:
+        log.exception(f"Error while queuing prompt: {e}")
+        raise e
+
+
+def get_image(filename, subfolder, folder_type, base_url):
+    log.info("get_image")
+    data = {"filename": filename, "subfolder": subfolder, "type": folder_type}
+    url_values = urllib.parse.urlencode(data)
+    req = urllib.request.Request(
+        f"{base_url}/view?{url_values}", headers=default_headers
+    )
+    with urllib.request.urlopen(req) as response:
+        return response.read()
+
+
+def get_image_url(filename, subfolder, folder_type, base_url):
+    log.info("get_image")
+    data = {"filename": filename, "subfolder": subfolder, "type": folder_type}
+    url_values = urllib.parse.urlencode(data)
+    return f"{base_url}/view?{url_values}"
+
+
+def get_history(prompt_id, base_url):
+    log.info("get_history")
+
+    req = urllib.request.Request(
+        f"{base_url}/history/{prompt_id}", headers=default_headers
+    )
+    with urllib.request.urlopen(req) as response:
+        return json.loads(response.read())
+
+
+def get_images(ws, prompt, client_id, base_url):
+    prompt_id = queue_prompt(prompt, client_id, base_url)["prompt_id"]
+    output_images = []
+    while True:
+        out = ws.recv()
+        if isinstance(out, str):
+            message = json.loads(out)
+            if message["type"] == "executing":
+                data = message["data"]
+                if data["node"] is None and data["prompt_id"] == prompt_id:
+                    break  # Execution is done
+        else:
+            continue  # previews are binary data
+
+    history = get_history(prompt_id, base_url)[prompt_id]
+    for o in history["outputs"]:
+        for node_id in history["outputs"]:
+            node_output = history["outputs"][node_id]
+            if "images" in node_output:
+                for image in node_output["images"]:
+                    url = get_image_url(
+                        image["filename"], image["subfolder"], image["type"], base_url
+                    )
+                    output_images.append({"url": url})
+    return {"data": output_images}
+
+
+class ComfyUINodeInput(BaseModel):
+    type: Optional[str] = None
+    node_ids: list[str] = []
+    key: Optional[str] = "text"
+    value: Optional[str] = None
+
+
+class ComfyUIWorkflow(BaseModel):
+    workflow: str
+    nodes: list[ComfyUINodeInput]
+
+
+class ComfyUIGenerateImageForm(BaseModel):
+    workflow: ComfyUIWorkflow
+
+    prompt: str
+    negative_prompt: Optional[str] = None
+    width: int
+    height: int
+    n: int = 1
+
+    steps: Optional[int] = None
+    seed: Optional[int] = None
+
+
+async def comfyui_generate_image(
+    model: str, payload: ComfyUIGenerateImageForm, client_id, base_url
+):
+    ws_url = base_url.replace("http://", "ws://").replace("https://", "wss://")
+    workflow = json.loads(payload.workflow.workflow)
+
+    for node in payload.workflow.nodes:
+        if node.type:
+            if node.type == "model":
+                for node_id in node.node_ids:
+                    workflow[node_id]["inputs"][node.key] = model
+            elif node.type == "prompt":
+                for node_id in node.node_ids:
+                    workflow[node_id]["inputs"]["text"] = payload.prompt
+            elif node.type == "negative_prompt":
+                for node_id in node.node_ids:
+                    workflow[node_id]["inputs"]["text"] = payload.negative_prompt
+            elif node.type == "width":
+                for node_id in node.node_ids:
+                    workflow[node_id]["inputs"]["width"] = payload.width
+            elif node.type == "height":
+                for node_id in node.node_ids:
+                    workflow[node_id]["inputs"]["height"] = payload.height
+            elif node.type == "n":
+                for node_id in node.node_ids:
+                    workflow[node_id]["inputs"]["batch_size"] = payload.n
+            elif node.type == "steps":
+                for node_id in node.node_ids:
+                    workflow[node_id]["inputs"]["steps"] = payload.steps
+            elif node.type == "seed":
+                seed = (
+                    payload.seed
+                    if payload.seed
+                    else random.randint(0, 18446744073709551614)
+                )
+                for node_id in node.node_ids:
+                    workflow[node_id]["inputs"][node.key] = seed
+        else:
+            for node_id in node.node_ids:
+                workflow[node_id]["inputs"][node.key] = node.value
+
+    try:
+        ws = websocket.WebSocket()
+        ws.connect(f"{ws_url}/ws?clientId={client_id}")
+        log.info("WebSocket connection established.")
+    except Exception as e:
+        log.exception(f"Failed to connect to WebSocket server: {e}")
+        return None
+
+    try:
+        log.info("Sending workflow to WebSocket server.")
+        log.info(f"Workflow: {workflow}")
+        images = await asyncio.to_thread(get_images, ws, workflow, client_id, base_url)
+    except Exception as e:
+        log.exception(f"Error while receiving images: {e}")
+        images = None
+
+    ws.close()
+
+    return images

backend/open_webui/apps/ollama/main.py

@@ -1,59 +1,44 @@
-from fastapi import (
-    FastAPI,
-    Request,
-    Response,
-    HTTPException,
-    Depends,
-    status,
-    UploadFile,
-    File,
-    BackgroundTasks,
-)
-from fastapi.middleware.cors import CORSMiddleware
-from fastapi.responses import StreamingResponse
-from fastapi.concurrency import run_in_threadpool
-
-from pydantic import BaseModel, ConfigDict
-
-import os
-import re
-import copy
-import random
-import requests
-import json
-import uuid
-import aiohttp
 import asyncio
+import json
 import logging
+import os
+import random
+import re
 import time
+from typing import Optional, Union
 from urllib.parse import urlparse
-from typing import Optional, List, Union
 
-from starlette.background import BackgroundTask
-
-from apps.webui.models.models import Models
-from apps.webui.models.users import Users
-from constants import ERROR_MESSAGES
-from utils.utils import (
-    decode_token,
-    get_current_user,
-    get_verified_user,
-    get_admin_user,
-)
-from utils.task import prompt_template
-
-
-from config import (
-    SRC_LOG_LEVELS,
-    OLLAMA_BASE_URLS,
-    ENABLE_OLLAMA_API,
+import aiohttp
+import requests
+from open_webui.apps.webui.models.models import Models
+from open_webui.config import (
     AIOHTTP_CLIENT_TIMEOUT,
+    CORS_ALLOW_ORIGIN,
     ENABLE_MODEL_FILTER,
+    ENABLE_OLLAMA_API,
     MODEL_FILTER_LIST,
+    OLLAMA_BASE_URLS,
     UPLOAD_DIR,
     AppConfig,
 )
-from utils.misc import calculate_sha256, add_or_update_system_message
+from open_webui.constants import ERROR_MESSAGES
+from open_webui.env import SRC_LOG_LEVELS
+from fastapi import Depends, FastAPI, File, HTTPException, Request, UploadFile
+from fastapi.middleware.cors import CORSMiddleware
+from fastapi.responses import StreamingResponse
+from pydantic import BaseModel, ConfigDict
+from starlette.background import BackgroundTask
+
+
+from open_webui.utils.misc import (
+    calculate_sha256,
+)
+from open_webui.utils.payload import (
+    apply_model_params_to_body_ollama,
+    apply_model_params_to_body_openai,
+    apply_model_system_prompt_to_body,
+)
+from open_webui.utils.utils import get_admin_user, get_verified_user
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["OLLAMA"])
@@ -61,7 +46,7 @@ log.setLevel(SRC_LOG_LEVELS["OLLAMA"])
 app = FastAPI()
 app.add_middleware(
     CORSMiddleware,
-    allow_origins=["*"],
+    allow_origins=CORS_ALLOW_ORIGIN,
     allow_credentials=True,
     allow_methods=["*"],
     allow_headers=["*"],
@@ -120,7 +105,7 @@ async def get_ollama_api_urls(user=Depends(get_admin_user)):
 
 
 class UrlUpdateForm(BaseModel):
-    urls: List[str]
+    urls: list[str]
 
 
 @app.post("/urls/update")
@@ -153,20 +138,29 @@ async def cleanup_response(
         await session.close()
 
 
-async def post_streaming_url(url: str, payload: str, stream: bool = True):
+async def post_streaming_url(
+    url: str, payload: Union[str, bytes], stream: bool = True, content_type=None
+):
     r = None
     try:
         session = aiohttp.ClientSession(
             trust_env=True, timeout=aiohttp.ClientTimeout(total=AIOHTTP_CLIENT_TIMEOUT)
         )
-        r = await session.post(url, data=payload)
+        r = await session.post(
+            url,
+            data=payload,
+            headers={"Content-Type": "application/json"},
+        )
         r.raise_for_status()
 
         if stream:
+            headers = dict(r.headers)
+            if content_type:
+                headers["Content-Type"] = content_type
             return StreamingResponse(
                 r.content,
                 status_code=r.status,
-                headers=dict(r.headers),
+                headers=headers,
                 background=BackgroundTask(
                     cleanup_response, response=r, session=session
                 ),
@@ -183,7 +177,7 @@ async def post_streaming_url(url: str, payload: str, stream: bool = True):
                 res = await r.json()
                 if "error" in res:
                     error_detail = f"Ollama: {res['error']}"
-            except:
+            except Exception:
                 error_detail = f"Ollama: {e}"
 
         raise HTTPException(
@@ -238,7 +232,7 @@ async def get_all_models():
 async def get_ollama_tags(
     url_idx: Optional[int] = None, user=Depends(get_verified_user)
 ):
-    if url_idx == None:
+    if url_idx is None:
         models = await get_all_models()
 
         if app.state.config.ENABLE_MODEL_FILTER:
@@ -269,7 +263,7 @@ async def get_ollama_tags(
                     res = r.json()
                     if "error" in res:
                         error_detail = f"Ollama: {res['error']}"
-                except:
+                except Exception:
                     error_detail = f"Ollama: {e}"
 
             raise HTTPException(
@@ -282,8 +276,7 @@ async def get_ollama_tags(
 @app.get("/api/version/{url_idx}")
 async def get_ollama_versions(url_idx: Optional[int] = None):
     if app.state.config.ENABLE_OLLAMA_API:
-        if url_idx == None:
-
+        if url_idx is None:
             # returns lowest version
             tasks = [
                 fetch_url(f"{url}/api/version")
@@ -323,7 +316,7 @@ async def get_ollama_versions(url_idx: Optional[int] = None):
                         res = r.json()
                         if "error" in res:
                             error_detail = f"Ollama: {res['error']}"
-                    except:
+                    except Exception:
                         error_detail = f"Ollama: {e}"
 
                 raise HTTPException(
@@ -346,8 +339,6 @@ async def pull_model(
     url = app.state.config.OLLAMA_BASE_URLS[url_idx]
     log.info(f"url: {url}")
 
-    r = None
-
     # Admin should be able to pull models from any source
     payload = {**form_data.model_dump(exclude_none=True), "insecure": True}
 
@@ -367,7 +358,7 @@ async def push_model(
     url_idx: Optional[int] = None,
     user=Depends(get_admin_user),
 ):
-    if url_idx == None:
+    if url_idx is None:
         if form_data.name in app.state.MODELS:
             url_idx = app.state.MODELS[form_data.name]["urls"][0]
         else:
@@ -417,7 +408,7 @@ async def copy_model(
     url_idx: Optional[int] = None,
     user=Depends(get_admin_user),
 ):
-    if url_idx == None:
+    if url_idx is None:
         if form_data.source in app.state.MODELS:
             url_idx = app.state.MODELS[form_data.source]["urls"][0]
         else:
@@ -428,13 +419,14 @@ async def copy_model(
 
     url = app.state.config.OLLAMA_BASE_URLS[url_idx]
     log.info(f"url: {url}")
+    r = requests.request(
+        method="POST",
+        url=f"{url}/api/copy",
+        headers={"Content-Type": "application/json"},
+        data=form_data.model_dump_json(exclude_none=True).encode(),
+    )
 
     try:
-        r = requests.request(
-            method="POST",
-            url=f"{url}/api/copy",
-            data=form_data.model_dump_json(exclude_none=True).encode(),
-        )
         r.raise_for_status()
 
         log.debug(f"r.text: {r.text}")
@@ -448,7 +440,7 @@ async def copy_model(
                 res = r.json()
                 if "error" in res:
                     error_detail = f"Ollama: {res['error']}"
-            except:
+            except Exception:
                 error_detail = f"Ollama: {e}"
 
         raise HTTPException(
@@ -464,7 +456,7 @@ async def delete_model(
     url_idx: Optional[int] = None,
     user=Depends(get_admin_user),
 ):
-    if url_idx == None:
+    if url_idx is None:
         if form_data.name in app.state.MODELS:
             url_idx = app.state.MODELS[form_data.name]["urls"][0]
         else:
@@ -476,12 +468,13 @@ async def delete_model(
     url = app.state.config.OLLAMA_BASE_URLS[url_idx]
     log.info(f"url: {url}")
 
+    r = requests.request(
+        method="DELETE",
+        url=f"{url}/api/delete",
+        headers={"Content-Type": "application/json"},
+        data=form_data.model_dump_json(exclude_none=True).encode(),
+    )
     try:
-        r = requests.request(
-            method="DELETE",
-            url=f"{url}/api/delete",
-            data=form_data.model_dump_json(exclude_none=True).encode(),
-        )
         r.raise_for_status()
 
         log.debug(f"r.text: {r.text}")
@@ -495,7 +488,7 @@ async def delete_model(
                 res = r.json()
                 if "error" in res:
                     error_detail = f"Ollama: {res['error']}"
-            except:
+            except Exception:
                 error_detail = f"Ollama: {e}"
 
         raise HTTPException(
@@ -516,12 +509,13 @@ async def show_model_info(form_data: ModelNameForm, user=Depends(get_verified_us
     url = app.state.config.OLLAMA_BASE_URLS[url_idx]
     log.info(f"url: {url}")
 
+    r = requests.request(
+        method="POST",
+        url=f"{url}/api/show",
+        headers={"Content-Type": "application/json"},
+        data=form_data.model_dump_json(exclude_none=True).encode(),
+    )
     try:
-        r = requests.request(
-            method="POST",
-            url=f"{url}/api/show",
-            data=form_data.model_dump_json(exclude_none=True).encode(),
-        )
         r.raise_for_status()
 
         return r.json()
@@ -533,7 +527,7 @@ async def show_model_info(form_data: ModelNameForm, user=Depends(get_verified_us
                 res = r.json()
                 if "error" in res:
                     error_detail = f"Ollama: {res['error']}"
-            except:
+            except Exception:
                 error_detail = f"Ollama: {e}"
 
         raise HTTPException(
@@ -549,14 +543,14 @@ class GenerateEmbeddingsForm(BaseModel):
     keep_alive: Optional[Union[int, str]] = None
 
 
-@app.post("/api/embeddings")
-@app.post("/api/embeddings/{url_idx}")
+@app.post("/api/embed")
+@app.post("/api/embed/{url_idx}")
 async def generate_embeddings(
     form_data: GenerateEmbeddingsForm,
     url_idx: Optional[int] = None,
     user=Depends(get_verified_user),
 ):
-    if url_idx == None:
+    if url_idx is None:
         model = form_data.model
 
         if ":" not in model:
@@ -573,12 +567,13 @@ async def generate_embeddings(
     url = app.state.config.OLLAMA_BASE_URLS[url_idx]
     log.info(f"url: {url}")
 
+    r = requests.request(
+        method="POST",
+        url=f"{url}/api/embed",
+        headers={"Content-Type": "application/json"},
+        data=form_data.model_dump_json(exclude_none=True).encode(),
+    )
     try:
-        r = requests.request(
-            method="POST",
-            url=f"{url}/api/embeddings",
-            data=form_data.model_dump_json(exclude_none=True).encode(),
-        )
         r.raise_for_status()
 
         return r.json()
@@ -590,7 +585,58 @@ async def generate_embeddings(
                 res = r.json()
                 if "error" in res:
                     error_detail = f"Ollama: {res['error']}"
-            except:
+            except Exception:
+                error_detail = f"Ollama: {e}"
+
+        raise HTTPException(
+            status_code=r.status_code if r else 500,
+            detail=error_detail,
+        )
+
+
+@app.post("/api/embeddings")
+@app.post("/api/embeddings/{url_idx}")
+async def generate_embeddings(
+    form_data: GenerateEmbeddingsForm,
+    url_idx: Optional[int] = None,
+    user=Depends(get_verified_user),
+):
+    if url_idx is None:
+        model = form_data.model
+
+        if ":" not in model:
+            model = f"{model}:latest"
+
+        if model in app.state.MODELS:
+            url_idx = random.choice(app.state.MODELS[model]["urls"])
+        else:
+            raise HTTPException(
+                status_code=400,
+                detail=ERROR_MESSAGES.MODEL_NOT_FOUND(form_data.model),
+            )
+
+    url = app.state.config.OLLAMA_BASE_URLS[url_idx]
+    log.info(f"url: {url}")
+
+    r = requests.request(
+        method="POST",
+        url=f"{url}/api/embeddings",
+        headers={"Content-Type": "application/json"},
+        data=form_data.model_dump_json(exclude_none=True).encode(),
+    )
+    try:
+        r.raise_for_status()
+
+        return r.json()
+    except Exception as e:
+        log.exception(e)
+        error_detail = "Open WebUI: Server Connection Error"
+        if r is not None:
+            try:
+                res = r.json()
+                if "error" in res:
+                    error_detail = f"Ollama: {res['error']}"
+            except Exception:
                 error_detail = f"Ollama: {e}"
 
         raise HTTPException(
@@ -603,10 +649,9 @@ def generate_ollama_embeddings(
     form_data: GenerateEmbeddingsForm,
     url_idx: Optional[int] = None,
 ):
-
     log.info(f"generate_ollama_embeddings {form_data}")
 
-    if url_idx == None:
+    if url_idx is None:
         model = form_data.model
 
         if ":" not in model:
@@ -623,12 +668,13 @@ def generate_ollama_embeddings(
     url = app.state.config.OLLAMA_BASE_URLS[url_idx]
     log.info(f"url: {url}")
 
+    r = requests.request(
+        method="POST",
+        url=f"{url}/api/embeddings",
+        headers={"Content-Type": "application/json"},
+        data=form_data.model_dump_json(exclude_none=True).encode(),
+    )
     try:
-        r = requests.request(
-            method="POST",
-            url=f"{url}/api/embeddings",
-            data=form_data.model_dump_json(exclude_none=True).encode(),
-        )
         r.raise_for_status()
 
         data = r.json()
@@ -638,7 +684,7 @@ def generate_ollama_embeddings(
         if "embedding" in data:
             return data["embedding"]
         else:
-            raise "Something went wrong :/"
+            raise Exception("Something went wrong :/")
     except Exception as e:
         log.exception(e)
         error_detail = "Open WebUI: Server Connection Error"
@@ -647,16 +693,16 @@ def generate_ollama_embeddings(
                 res = r.json()
                 if "error" in res:
                     error_detail = f"Ollama: {res['error']}"
-            except:
+            except Exception:
                 error_detail = f"Ollama: {e}"
 
-        raise error_detail
+        raise Exception(error_detail)
 
 
 class GenerateCompletionForm(BaseModel):
     model: str
     prompt: str
-    images: Optional[List[str]] = None
+    images: Optional[list[str]] = None
     format: Optional[str] = None
     options: Optional[dict] = None
     system: Optional[str] = None
@@ -674,8 +720,7 @@ async def generate_completion(
     url_idx: Optional[int] = None,
     user=Depends(get_verified_user),
 ):
-
-    if url_idx == None:
+    if url_idx is None:
         model = form_data.model
 
         if ":" not in model:
@@ -700,12 +745,12 @@ async def generate_completion(
 class ChatMessage(BaseModel):
     role: str
     content: str
-    images: Optional[List[str]] = None
+    images: Optional[list[str]] = None
 
 
 class GenerateChatCompletionForm(BaseModel):
     model: str
-    messages: List[ChatMessage]
+    messages: list[ChatMessage]
     format: Optional[str] = None
     options: Optional[dict] = None
     template: Optional[str] = None
@@ -713,6 +758,18 @@ class GenerateChatCompletionForm(BaseModel):
     keep_alive: Optional[Union[int, str]] = None
 
 
+def get_ollama_url(url_idx: Optional[int], model: str):
+    if url_idx is None:
+        if model not in app.state.MODELS:
+            raise HTTPException(
+                status_code=400,
+                detail=ERROR_MESSAGES.MODEL_NOT_FOUND(model),
+            )
+        url_idx = random.choice(app.state.MODELS[model]["urls"])
+    url = app.state.config.OLLAMA_BASE_URLS[url_idx]
+    return url
+
+
 @app.post("/api/chat")
 @app.post("/api/chat/{url_idx}")
 async def generate_chat_completion(
@@ -720,209 +777,50 @@ async def generate_chat_completion(
     url_idx: Optional[int] = None,
     user=Depends(get_verified_user),
 ):
-
-    log.debug(
-        "form_data.model_dump_json(exclude_none=True).encode(): {0} ".format(
-            form_data.model_dump_json(exclude_none=True).encode()
-        )
-    )
-
-    payload = {
-        **form_data.model_dump(exclude_none=True, exclude=["metadata"]),
-    }
+    payload = {**form_data.model_dump(exclude_none=True)}
+    log.debug(f"{payload = }")
     if "metadata" in payload:
         del payload["metadata"]
 
     model_id = form_data.model
+
+    if app.state.config.ENABLE_MODEL_FILTER:
+        if user.role == "user" and model_id not in app.state.config.MODEL_FILTER_LIST:
+            raise HTTPException(
+                status_code=403,
+                detail="Model not found",
+            )
+
     model_info = Models.get_model_by_id(model_id)
 
     if model_info:
         if model_info.base_model_id:
             payload["model"] = model_info.base_model_id
 
-        model_info.params = model_info.params.model_dump()
+        params = model_info.params.model_dump()
 
-        if model_info.params:
+        if params:
             if payload.get("options") is None:
                 payload["options"] = {}
 
-            if (
-                model_info.params.get("mirostat", None)
-                and payload["options"].get("mirostat") is None
-            ):
-                payload["options"]["mirostat"] = model_info.params.get("mirostat", None)
-
-            if (
-                model_info.params.get("mirostat_eta", None)
-                and payload["options"].get("mirostat_eta") is None
-            ):
-                payload["options"]["mirostat_eta"] = model_info.params.get(
-                    "mirostat_eta", None
-                )
-
-            if (
-                model_info.params.get("mirostat_tau", None)
-                and payload["options"].get("mirostat_tau") is None
-            ):
-                payload["options"]["mirostat_tau"] = model_info.params.get(
-                    "mirostat_tau", None
-                )
-
-            if (
-                model_info.params.get("num_ctx", None)
-                and payload["options"].get("num_ctx") is None
-            ):
-                payload["options"]["num_ctx"] = model_info.params.get("num_ctx", None)
-
-            if (
-                model_info.params.get("num_batch", None)
-                and payload["options"].get("num_batch") is None
-            ):
-                payload["options"]["num_batch"] = model_info.params.get(
-                    "num_batch", None
-                )
-
-            if (
-                model_info.params.get("num_keep", None)
-                and payload["options"].get("num_keep") is None
-            ):
-                payload["options"]["num_keep"] = model_info.params.get("num_keep", None)
-
-            if (
-                model_info.params.get("repeat_last_n", None)
-                and payload["options"].get("repeat_last_n") is None
-            ):
-                payload["options"]["repeat_last_n"] = model_info.params.get(
-                    "repeat_last_n", None
-                )
-
-            if (
-                model_info.params.get("frequency_penalty", None)
-                and payload["options"].get("frequency_penalty") is None
-            ):
-                payload["options"]["repeat_penalty"] = model_info.params.get(
-                    "frequency_penalty", None
-                )
-
-            if (
-                model_info.params.get("temperature", None) is not None
-                and payload["options"].get("temperature") is None
-            ):
-                payload["options"]["temperature"] = model_info.params.get(
-                    "temperature", None
-                )
-
-            if (
-                model_info.params.get("seed", None) is not None
-                and payload["options"].get("seed") is None
-            ):
-                payload["options"]["seed"] = model_info.params.get("seed", None)
-
-            if (
-                model_info.params.get("stop", None)
-                and payload["options"].get("stop") is None
-            ):
-                payload["options"]["stop"] = (
-                    [
-                        bytes(stop, "utf-8").decode("unicode_escape")
-                        for stop in model_info.params["stop"]
-                    ]
-                    if model_info.params.get("stop", None)
-                    else None
-                )
-
-            if (
-                model_info.params.get("tfs_z", None)
-                and payload["options"].get("tfs_z") is None
-            ):
-                payload["options"]["tfs_z"] = model_info.params.get("tfs_z", None)
-
-            if (
-                model_info.params.get("max_tokens", None)
-                and payload["options"].get("max_tokens") is None
-            ):
-                payload["options"]["num_predict"] = model_info.params.get(
-                    "max_tokens", None
-                )
-
-            if (
-                model_info.params.get("top_k", None)
-                and payload["options"].get("top_k") is None
-            ):
-                payload["options"]["top_k"] = model_info.params.get("top_k", None)
-
-            if (
-                model_info.params.get("top_p", None)
-                and payload["options"].get("top_p") is None
-            ):
-                payload["options"]["top_p"] = model_info.params.get("top_p", None)
-
-            if (
-                model_info.params.get("min_p", None)
-                and payload["options"].get("min_p") is None
-            ):
-                payload["options"]["min_p"] = model_info.params.get("min_p", None)
-
-            if (
-                model_info.params.get("use_mmap", None)
-                and payload["options"].get("use_mmap") is None
-            ):
-                payload["options"]["use_mmap"] = model_info.params.get("use_mmap", None)
-
-            if (
-                model_info.params.get("use_mlock", None)
-                and payload["options"].get("use_mlock") is None
-            ):
-                payload["options"]["use_mlock"] = model_info.params.get(
-                    "use_mlock", None
-                )
-
-            if (
-                model_info.params.get("num_thread", None)
-                and payload["options"].get("num_thread") is None
-            ):
-                payload["options"]["num_thread"] = model_info.params.get(
-                    "num_thread", None
-                )
-
-        system = model_info.params.get("system", None)
-        if system:
-            system = prompt_template(
-                system,
-                **(
-                    {
-                        "user_name": user.name,
-                        "user_location": (
-                            user.info.get("location") if user.info else None
-                        ),
-                    }
-                    if user
-                    else {}
-                ),
+            payload["options"] = apply_model_params_to_body_ollama(
+                params, payload["options"]
             )
+            payload = apply_model_system_prompt_to_body(params, payload, user)
 
-            if payload.get("messages"):
-                payload["messages"] = add_or_update_system_message(
-                    system, payload["messages"]
-                )
+    if ":" not in payload["model"]:
+        payload["model"] = f"{payload['model']}:latest"
 
-    if url_idx == None:
-        if ":" not in payload["model"]:
-            payload["model"] = f"{payload['model']}:latest"
-
-        if payload["model"] in app.state.MODELS:
-            url_idx = random.choice(app.state.MODELS[payload["model"]]["urls"])
-        else:
-            raise HTTPException(
-                status_code=400,
-                detail=ERROR_MESSAGES.MODEL_NOT_FOUND(form_data.model),
-            )
-
-    url = app.state.config.OLLAMA_BASE_URLS[url_idx]
+    url = get_ollama_url(url_idx, payload["model"])
     log.info(f"url: {url}")
     log.debug(payload)
 
-    return await post_streaming_url(f"{url}/api/chat", json.dumps(payload))
+    return await post_streaming_url(
+        f"{url}/api/chat",
+        json.dumps(payload),
+        stream=form_data.stream,
+        content_type="application/x-ndjson",
+    )
 
 
 # TODO: we should update this part once Ollama supports other types
@@ -940,7 +838,7 @@ class OpenAIChatMessage(BaseModel):
 
 class OpenAIChatCompletionForm(BaseModel):
     model: str
-    messages: List[OpenAIChatMessage]
+    messages: list[OpenAIChatMessage]
 
     model_config = ConfigDict(extra="allow")
 
@@ -952,83 +850,36 @@ async def generate_openai_chat_completion(
     url_idx: Optional[int] = None,
     user=Depends(get_verified_user),
 ):
-    form_data = OpenAIChatCompletionForm(**form_data)
-    payload = {**form_data.model_dump(exclude_none=True, exclude=["metadata"])}
-
+    completion_form = OpenAIChatCompletionForm(**form_data)
+    payload = {**completion_form.model_dump(exclude_none=True, exclude=["metadata"])}
     if "metadata" in payload:
         del payload["metadata"]
 
-    model_id = form_data.model
+    model_id = completion_form.model
+
+    if app.state.config.ENABLE_MODEL_FILTER:
+        if user.role == "user" and model_id not in app.state.config.MODEL_FILTER_LIST:
+            raise HTTPException(
+                status_code=403,
+                detail="Model not found",
+            )
+
     model_info = Models.get_model_by_id(model_id)
 
     if model_info:
         if model_info.base_model_id:
             payload["model"] = model_info.base_model_id
 
-        model_info.params = model_info.params.model_dump()
+        params = model_info.params.model_dump()
 
-        if model_info.params:
-            payload["temperature"] = model_info.params.get("temperature", None)
-            payload["top_p"] = model_info.params.get("top_p", None)
-            payload["max_tokens"] = model_info.params.get("max_tokens", None)
-            payload["frequency_penalty"] = model_info.params.get(
-                "frequency_penalty", None
-            )
-            payload["seed"] = model_info.params.get("seed", None)
-            payload["stop"] = (
-                [
-                    bytes(stop, "utf-8").decode("unicode_escape")
-                    for stop in model_info.params["stop"]
-                ]
-                if model_info.params.get("stop", None)
-                else None
-            )
+        if params:
+            payload = apply_model_params_to_body_openai(params, payload)
+            payload = apply_model_system_prompt_to_body(params, payload, user)
 
-        system = model_info.params.get("system", None)
+    if ":" not in payload["model"]:
+        payload["model"] = f"{payload['model']}:latest"
 
-        if system:
-            system = prompt_template(
-                system,
-                **(
-                    {
-                        "user_name": user.name,
-                        "user_location": (
-                            user.info.get("location") if user.info else None
-                        ),
-                    }
-                    if user
-                    else {}
-                ),
-            )
-            # Check if the payload already has a system message
-            # If not, add a system message to the payload
-            if payload.get("messages"):
-                for message in payload["messages"]:
-                    if message.get("role") == "system":
-                        message["content"] = system + message["content"]
-                        break
-                else:
-                    payload["messages"].insert(
-                        0,
-                        {
-                            "role": "system",
-                            "content": system,
-                        },
-                    )
-
-    if url_idx == None:
-        if ":" not in payload["model"]:
-            payload["model"] = f"{payload['model']}:latest"
-
-        if payload["model"] in app.state.MODELS:
-            url_idx = random.choice(app.state.MODELS[payload["model"]]["urls"])
-        else:
-            raise HTTPException(
-                status_code=400,
-                detail=ERROR_MESSAGES.MODEL_NOT_FOUND(form_data.model),
-            )
-
-    url = app.state.config.OLLAMA_BASE_URLS[url_idx]
+    url = get_ollama_url(url_idx, payload["model"])
     log.info(f"url: {url}")
 
     return await post_streaming_url(
@@ -1044,7 +895,7 @@ async def get_openai_models(
     url_idx: Optional[int] = None,
     user=Depends(get_verified_user),
 ):
-    if url_idx == None:
+    if url_idx is None:
         models = await get_all_models()
 
         if app.state.config.ENABLE_MODEL_FILTER:
@@ -1099,7 +950,7 @@ async def get_openai_models(
                     res = r.json()
                     if "error" in res:
                         error_detail = f"Ollama: {res['error']}"
-                except:
+                except Exception:
                     error_detail = f"Ollama: {e}"
 
             raise HTTPException(
@@ -1125,7 +976,6 @@ def parse_huggingface_url(hf_url):
         path_components = parsed_url.path.split("/")
 
         # Extract the desired output
-        user_repo = "/".join(path_components[1:3])
         model_file = path_components[-1]
 
         return model_file
@@ -1190,7 +1040,6 @@ async def download_model(
     url_idx: Optional[int] = None,
     user=Depends(get_admin_user),
 ):
-
     allowed_hosts = ["https://huggingface.co/", "https://github.com/"]
 
     if not any(form_data.url.startswith(host) for host in allowed_hosts):
@@ -1199,7 +1048,7 @@ async def download_model(
             detail="Invalid file_url. Only URLs from allowed hosts are permitted.",
         )
 
-    if url_idx == None:
+    if url_idx is None:
         url_idx = 0
     url = app.state.config.OLLAMA_BASE_URLS[url_idx]
 
@@ -1222,7 +1071,7 @@ def upload_model(
     url_idx: Optional[int] = None,
     user=Depends(get_admin_user),
 ):
-    if url_idx == None:
+    if url_idx is None:
         url_idx = 0
     ollama_url = app.state.config.OLLAMA_BASE_URLS[url_idx]
 

backend/open_webui/apps/openai/main.py

@@ -1,40 +1,39 @@
-from fastapi import FastAPI, Request, HTTPException, Depends
-from fastapi.middleware.cors import CORSMiddleware
-from fastapi.responses import StreamingResponse, FileResponse
-
-import requests
-import aiohttp
 import asyncio
+import hashlib
 import json
 import logging
+from pathlib import Path
+from typing import Literal, Optional, overload
 
+import aiohttp
+import requests
+from open_webui.apps.webui.models.models import Models
+from open_webui.config import (
+    AIOHTTP_CLIENT_TIMEOUT,
+    CACHE_DIR,
+    CORS_ALLOW_ORIGIN,
+    ENABLE_MODEL_FILTER,
+    ENABLE_OPENAI_API,
+    MODEL_FILTER_LIST,
+    OPENAI_API_BASE_URLS,
+    OPENAI_API_KEYS,
+    AppConfig,
+)
+from open_webui.constants import ERROR_MESSAGES
+from open_webui.env import SRC_LOG_LEVELS
+from fastapi import Depends, FastAPI, HTTPException, Request
+from fastapi.middleware.cors import CORSMiddleware
+from fastapi.responses import FileResponse, StreamingResponse
 from pydantic import BaseModel
 from starlette.background import BackgroundTask
 
-from apps.webui.models.models import Models
-from constants import ERROR_MESSAGES
-from utils.utils import (
-    get_verified_user,
-    get_admin_user,
+
+from open_webui.utils.payload import (
+    apply_model_params_to_body_openai,
+    apply_model_system_prompt_to_body,
 )
-from utils.misc import apply_model_params_to_body, apply_model_system_prompt_to_body
 
-from config import (
-    SRC_LOG_LEVELS,
-    ENABLE_OPENAI_API,
-    AIOHTTP_CLIENT_TIMEOUT,
-    OPENAI_API_BASE_URLS,
-    OPENAI_API_KEYS,
-    CACHE_DIR,
-    ENABLE_MODEL_FILTER,
-    MODEL_FILTER_LIST,
-    AppConfig,
-)
-from typing import List, Optional, Literal, overload
-
-
-import hashlib
-from pathlib import Path
+from open_webui.utils.utils import get_admin_user, get_verified_user
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["OPENAI"])
@@ -42,7 +41,7 @@ log.setLevel(SRC_LOG_LEVELS["OPENAI"])
 app = FastAPI()
 app.add_middleware(
     CORSMiddleware,
-    allow_origins=["*"],
+    allow_origins=CORS_ALLOW_ORIGIN,
     allow_credentials=True,
     allow_methods=["*"],
     allow_headers=["*"],
@@ -86,11 +85,11 @@ async def update_config(form_data: OpenAIConfigForm, user=Depends(get_admin_user
 
 
 class UrlsUpdateForm(BaseModel):
-    urls: List[str]
+    urls: list[str]
 
 
 class KeysUpdateForm(BaseModel):
-    keys: List[str]
+    keys: list[str]
 
 
 @app.get("/urls")
@@ -221,7 +220,17 @@ def merge_models_lists(model_lists):
                     for model in models
                     if "api.openai.com"
                     not in app.state.config.OPENAI_API_BASE_URLS[idx]
-                    or "gpt" in model["id"]
+                    or not any(
+                        name in model["id"]
+                        for name in [
+                            "babbage",
+                            "dall-e",
+                            "davinci",
+                            "embedding",
+                            "tts",
+                            "whisper",
+                        ]
+                    )
                 ]
             )
 
@@ -324,10 +333,24 @@ async def get_models(url_idx: Optional[int] = None, user=Depends(get_verified_us
             r.raise_for_status()
 
             response_data = r.json()
+
             if "api.openai.com" in url:
-                response_data["data"] = list(
-                    filter(lambda model: "gpt" in model["id"], response_data["data"])
-                )
+                # Filter the response data
+                response_data["data"] = [
+                    model
+                    for model in response_data["data"]
+                    if not any(
+                        name in model["id"]
+                        for name in [
+                            "babbage",
+                            "dall-e",
+                            "davinci",
+                            "embedding",
+                            "tts",
+                            "whisper",
+                        ]
+                    )
+                ]
 
             return response_data
         except Exception as e:
@@ -356,7 +379,9 @@ async def generate_chat_completion(
 ):
     idx = 0
     payload = {**form_data}
-    payload.pop("metadata")
+
+    if "metadata" in payload:
+        del payload["metadata"]
 
     model_id = form_data.get("model")
     model_info = Models.get_model_by_id(model_id)
@@ -366,7 +391,7 @@ async def generate_chat_completion(
             payload["model"] = model_info.base_model_id
 
         params = model_info.params.model_dump()
-        payload = apply_model_params_to_body(params, payload)
+        payload = apply_model_params_to_body_openai(params, payload)
         payload = apply_model_system_prompt_to_body(params, payload, user)
 
     model = app.state.MODELS[payload.get("model")]
@@ -380,14 +405,19 @@ async def generate_chat_completion(
             "role": user.role,
         }
 
+    url = app.state.config.OPENAI_API_BASE_URLS[idx]
+    key = app.state.config.OPENAI_API_KEYS[idx]
+
+    # Change max_completion_tokens to max_tokens (Backward compatible)
+    if "api.openai.com" not in url and not payload["model"].lower().startswith("o1-"):
+        if "max_completion_tokens" in payload:
+            payload["max_tokens"] = payload.pop("max_completion_tokens")
+
     # Convert the modified body back to JSON
     payload = json.dumps(payload)
 
     log.debug(payload)
 
-    url = app.state.config.OPENAI_API_BASE_URLS[idx]
-    key = app.state.config.OPENAI_API_KEYS[idx]
-
     headers = {}
     headers["Authorization"] = f"Bearer {key}"
     headers["Content-Type"] = "application/json"
@@ -398,6 +428,7 @@ async def generate_chat_completion(
     r = None
     session = None
     streaming = False
+    response = None
 
     try:
         session = aiohttp.ClientSession(
@@ -410,8 +441,6 @@ async def generate_chat_completion(
             headers=headers,
         )
 
-        r.raise_for_status()
-
         # Check if response is SSE
         if "text/event-stream" in r.headers.get("Content-Type", ""):
             streaming = True
@@ -424,19 +453,23 @@ async def generate_chat_completion(
                 ),
             )
         else:
-            response_data = await r.json()
-            return response_data
+            try:
+                response = await r.json()
+            except Exception as e:
+                log.error(e)
+                response = await r.text()
+
+            r.raise_for_status()
+            return response
     except Exception as e:
         log.exception(e)
         error_detail = "Open WebUI: Server Connection Error"
-        if r is not None:
-            try:
-                res = await r.json()
-                print(res)
-                if "error" in res:
-                    error_detail = f"External: {res['error']['message'] if 'message' in res['error'] else res['error']}"
-            except Exception:
-                error_detail = f"External: {e}"
+        if isinstance(response, dict):
+            if "error" in response:
+                error_detail = f"{response['error']['message'] if 'message' in response['error'] else response['error']}"
+        elif isinstance(response, str):
+            error_detail = response
+
         raise HTTPException(status_code=r.status if r else 500, detail=error_detail)
     finally:
         if not streaming and session:

backend/open_webui/apps/rag/main.py

@@ -1,137 +1,125 @@
-from fastapi import (
-    FastAPI,
-    Depends,
-    HTTPException,
-    status,
-    UploadFile,
-    File,
-    Form,
-)
-from fastapi.middleware.cors import CORSMiddleware
-import requests
-import os, shutil, logging, re
-from datetime import datetime
-
-from pathlib import Path
-from typing import List, Union, Sequence, Iterator, Any
-
-from chromadb.utils.batch_utils import create_batches
-from langchain_core.documents import Document
-
-from langchain_community.document_loaders import (
-    WebBaseLoader,
-    TextLoader,
-    PyPDFLoader,
-    CSVLoader,
-    BSHTMLLoader,
-    Docx2txtLoader,
-    UnstructuredEPubLoader,
-    UnstructuredWordDocumentLoader,
-    UnstructuredMarkdownLoader,
-    UnstructuredXMLLoader,
-    UnstructuredRSTLoader,
-    UnstructuredExcelLoader,
-    UnstructuredPowerPointLoader,
-    YoutubeLoader,
-    OutlookMessageLoader,
-)
-from langchain.text_splitter import RecursiveCharacterTextSplitter
-
-import validators
-import urllib.parse
-import socket
-
-
-from pydantic import BaseModel
-from typing import Optional
-import mimetypes
-import uuid
 import json
+import logging
+import mimetypes
+import os
+import shutil
+import socket
+import urllib.parse
+import uuid
+from datetime import datetime
+from pathlib import Path
+from typing import Iterator, Optional, Sequence, Union
 
-from apps.webui.models.documents import (
-    Documents,
-    DocumentForm,
-    DocumentResponse,
-)
-from apps.webui.models.files import (
-    Files,
-)
 
-from apps.rag.utils import (
-    get_model_path,
+import numpy as np
+import torch
+import requests
+import validators
+
+from fastapi import Depends, FastAPI, File, Form, HTTPException, UploadFile, status
+from fastapi.middleware.cors import CORSMiddleware
+from pydantic import BaseModel
+
+from open_webui.apps.rag.search.main import SearchResult
+from open_webui.apps.rag.search.brave import search_brave
+from open_webui.apps.rag.search.duckduckgo import search_duckduckgo
+from open_webui.apps.rag.search.google_pse import search_google_pse
+from open_webui.apps.rag.search.jina_search import search_jina
+from open_webui.apps.rag.search.searchapi import search_searchapi
+from open_webui.apps.rag.search.searxng import search_searxng
+from open_webui.apps.rag.search.serper import search_serper
+from open_webui.apps.rag.search.serply import search_serply
+from open_webui.apps.rag.search.serpstack import search_serpstack
+from open_webui.apps.rag.search.tavily import search_tavily
+from open_webui.apps.rag.utils import (
     get_embedding_function,
-    query_doc,
-    query_doc_with_hybrid_search,
+    get_model_path,
     query_collection,
     query_collection_with_hybrid_search,
+    query_doc,
+    query_doc_with_hybrid_search,
 )
-
-from apps.rag.search.brave import search_brave
-from apps.rag.search.google_pse import search_google_pse
-from apps.rag.search.main import SearchResult
-from apps.rag.search.searxng import search_searxng
-from apps.rag.search.serper import search_serper
-from apps.rag.search.serpstack import search_serpstack
-from apps.rag.search.serply import search_serply
-from apps.rag.search.duckduckgo import search_duckduckgo
-from apps.rag.search.tavily import search_tavily
-from apps.rag.search.jina_search import search_jina
-
-from utils.misc import (
-    calculate_sha256,
-    calculate_sha256_string,
-    sanitize_filename,
-    extract_folders_after_data_docs,
-)
-from utils.utils import get_verified_user, get_admin_user
-
-from config import (
-    AppConfig,
-    ENV,
-    SRC_LOG_LEVELS,
-    UPLOAD_DIR,
-    DOCS_DIR,
+from open_webui.apps.webui.models.documents import DocumentForm, Documents
+from open_webui.apps.webui.models.files import Files
+from open_webui.config import (
+    BRAVE_SEARCH_API_KEY,
+    CHUNK_OVERLAP,
+    CHUNK_SIZE,
     CONTENT_EXTRACTION_ENGINE,
-    TIKA_SERVER_URL,
-    RAG_TOP_K,
-    RAG_RELEVANCE_THRESHOLD,
+    CORS_ALLOW_ORIGIN,
+    DOCS_DIR,
+    ENABLE_RAG_HYBRID_SEARCH,
+    ENABLE_RAG_LOCAL_WEB_FETCH,
+    ENABLE_RAG_WEB_LOADER_SSL_VERIFICATION,
+    ENABLE_RAG_WEB_SEARCH,
+    ENV,
+    GOOGLE_PSE_API_KEY,
+    GOOGLE_PSE_ENGINE_ID,
+    PDF_EXTRACT_IMAGES,
     RAG_EMBEDDING_ENGINE,
     RAG_EMBEDDING_MODEL,
     RAG_EMBEDDING_MODEL_AUTO_UPDATE,
     RAG_EMBEDDING_MODEL_TRUST_REMOTE_CODE,
-    ENABLE_RAG_HYBRID_SEARCH,
-    ENABLE_RAG_WEB_LOADER_SSL_VERIFICATION,
-    RAG_RERANKING_MODEL,
-    PDF_EXTRACT_IMAGES,
-    RAG_RERANKING_MODEL_AUTO_UPDATE,
-    RAG_RERANKING_MODEL_TRUST_REMOTE_CODE,
+    RAG_EMBEDDING_OPENAI_BATCH_SIZE,
+    RAG_FILE_MAX_COUNT,
+    RAG_FILE_MAX_SIZE,
     RAG_OPENAI_API_BASE_URL,
     RAG_OPENAI_API_KEY,
-    DEVICE_TYPE,
-    CHROMA_CLIENT,
-    CHUNK_SIZE,
-    CHUNK_OVERLAP,
+    RAG_RELEVANCE_THRESHOLD,
+    RAG_RERANKING_MODEL,
+    RAG_RERANKING_MODEL_AUTO_UPDATE,
+    RAG_RERANKING_MODEL_TRUST_REMOTE_CODE,
+    DEFAULT_RAG_TEMPLATE,
     RAG_TEMPLATE,
-    ENABLE_RAG_LOCAL_WEB_FETCH,
-    YOUTUBE_LOADER_LANGUAGE,
-    ENABLE_RAG_WEB_SEARCH,
-    RAG_WEB_SEARCH_ENGINE,
+    RAG_TOP_K,
+    RAG_WEB_SEARCH_CONCURRENT_REQUESTS,
     RAG_WEB_SEARCH_DOMAIN_FILTER_LIST,
+    RAG_WEB_SEARCH_ENGINE,
+    RAG_WEB_SEARCH_RESULT_COUNT,
+    SEARCHAPI_API_KEY,
+    SEARCHAPI_ENGINE,
     SEARXNG_QUERY_URL,
-    GOOGLE_PSE_API_KEY,
-    GOOGLE_PSE_ENGINE_ID,
-    BRAVE_SEARCH_API_KEY,
-    SERPSTACK_API_KEY,
-    SERPSTACK_HTTPS,
     SERPER_API_KEY,
     SERPLY_API_KEY,
+    SERPSTACK_API_KEY,
+    SERPSTACK_HTTPS,
     TAVILY_API_KEY,
-    RAG_WEB_SEARCH_RESULT_COUNT,
-    RAG_WEB_SEARCH_CONCURRENT_REQUESTS,
-    RAG_EMBEDDING_OPENAI_BATCH_SIZE,
+    TIKA_SERVER_URL,
+    UPLOAD_DIR,
+    YOUTUBE_LOADER_LANGUAGE,
+    AppConfig,
 )
+from open_webui.constants import ERROR_MESSAGES
+from open_webui.env import SRC_LOG_LEVELS, DEVICE_TYPE, DOCKER
+from open_webui.utils.misc import (
+    calculate_sha256,
+    calculate_sha256_string,
+    extract_folders_after_data_docs,
+    sanitize_filename,
+)
+from open_webui.utils.utils import get_admin_user, get_verified_user
+from open_webui.apps.rag.vector.connector import VECTOR_DB_CLIENT
 
-from constants import ERROR_MESSAGES
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+from langchain_community.document_loaders import (
+    BSHTMLLoader,
+    CSVLoader,
+    Docx2txtLoader,
+    OutlookMessageLoader,
+    PyPDFLoader,
+    TextLoader,
+    UnstructuredEPubLoader,
+    UnstructuredExcelLoader,
+    UnstructuredMarkdownLoader,
+    UnstructuredPowerPointLoader,
+    UnstructuredRSTLoader,
+    UnstructuredXMLLoader,
+    WebBaseLoader,
+    YoutubeLoader,
+)
+from langchain_core.documents import Document
+from colbert.infra import ColBERTConfig
+from colbert.modeling.checkpoint import Checkpoint
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["RAG"])
@@ -142,6 +130,8 @@ app.state.config = AppConfig()
 
 app.state.config.TOP_K = RAG_TOP_K
 app.state.config.RELEVANCE_THRESHOLD = RAG_RELEVANCE_THRESHOLD
+app.state.config.FILE_MAX_SIZE = RAG_FILE_MAX_SIZE
+app.state.config.FILE_MAX_COUNT = RAG_FILE_MAX_COUNT
 
 app.state.config.ENABLE_RAG_HYBRID_SEARCH = ENABLE_RAG_HYBRID_SEARCH
 app.state.config.ENABLE_RAG_WEB_LOADER_SSL_VERIFICATION = (
@@ -160,13 +150,11 @@ app.state.config.RAG_EMBEDDING_OPENAI_BATCH_SIZE = RAG_EMBEDDING_OPENAI_BATCH_SI
 app.state.config.RAG_RERANKING_MODEL = RAG_RERANKING_MODEL
 app.state.config.RAG_TEMPLATE = RAG_TEMPLATE
 
-
 app.state.config.OPENAI_API_BASE_URL = RAG_OPENAI_API_BASE_URL
 app.state.config.OPENAI_API_KEY = RAG_OPENAI_API_KEY
 
 app.state.config.PDF_EXTRACT_IMAGES = PDF_EXTRACT_IMAGES
 
-
 app.state.config.YOUTUBE_LOADER_LANGUAGE = YOUTUBE_LOADER_LANGUAGE
 app.state.YOUTUBE_LOADER_TRANSLATION = None
 
@@ -184,19 +172,21 @@ app.state.config.SERPSTACK_HTTPS = SERPSTACK_HTTPS
 app.state.config.SERPER_API_KEY = SERPER_API_KEY
 app.state.config.SERPLY_API_KEY = SERPLY_API_KEY
 app.state.config.TAVILY_API_KEY = TAVILY_API_KEY
+app.state.config.SEARCHAPI_API_KEY = SEARCHAPI_API_KEY
+app.state.config.SEARCHAPI_ENGINE = SEARCHAPI_ENGINE
 app.state.config.RAG_WEB_SEARCH_RESULT_COUNT = RAG_WEB_SEARCH_RESULT_COUNT
 app.state.config.RAG_WEB_SEARCH_CONCURRENT_REQUESTS = RAG_WEB_SEARCH_CONCURRENT_REQUESTS
 
 
 def update_embedding_model(
     embedding_model: str,
-    update_model: bool = False,
+    auto_update: bool = False,
 ):
     if embedding_model and app.state.config.RAG_EMBEDDING_ENGINE == "":
         import sentence_transformers
 
         app.state.sentence_transformer_ef = sentence_transformers.SentenceTransformer(
-            get_model_path(embedding_model, update_model),
+            get_model_path(embedding_model, auto_update),
             device=DEVICE_TYPE,
             trust_remote_code=RAG_EMBEDDING_MODEL_TRUST_REMOTE_CODE,
         )
@@ -206,16 +196,108 @@ def update_embedding_model(
 
 def update_reranking_model(
     reranking_model: str,
-    update_model: bool = False,
+    auto_update: bool = False,
 ):
     if reranking_model:
-        import sentence_transformers
+        if any(model in reranking_model for model in ["jinaai/jina-colbert-v2"]):
 
-        app.state.sentence_transformer_rf = sentence_transformers.CrossEncoder(
-            get_model_path(reranking_model, update_model),
-            device=DEVICE_TYPE,
-            trust_remote_code=RAG_RERANKING_MODEL_TRUST_REMOTE_CODE,
-        )
+            class ColBERT:
+                def __init__(self, name) -> None:
+                    print("ColBERT: Loading model", name)
+                    self.device = "cuda" if torch.cuda.is_available() else "cpu"
+
+                    if DOCKER:
+                        # This is a workaround for the issue with the docker container
+                        # where the torch extension is not loaded properly
+                        # and the following error is thrown:
+                        # /root/.cache/torch_extensions/py311_cpu/segmented_maxsim_cpp/segmented_maxsim_cpp.so: cannot open shared object file: No such file or directory
+
+                        lock_file = "/root/.cache/torch_extensions/py311_cpu/segmented_maxsim_cpp/lock"
+                        if os.path.exists(lock_file):
+                            os.remove(lock_file)
+
+                    self.ckpt = Checkpoint(
+                        name,
+                        colbert_config=ColBERTConfig(model_name=name),
+                    ).to(self.device)
+                    pass
+
+                def calculate_similarity_scores(
+                    self, query_embeddings, document_embeddings
+                ):
+
+                    query_embeddings = query_embeddings.to(self.device)
+                    document_embeddings = document_embeddings.to(self.device)
+
+                    # Validate dimensions to ensure compatibility
+                    if query_embeddings.dim() != 3:
+                        raise ValueError(
+                            f"Expected query embeddings to have 3 dimensions, but got {query_embeddings.dim()}."
+                        )
+                    if document_embeddings.dim() != 3:
+                        raise ValueError(
+                            f"Expected document embeddings to have 3 dimensions, but got {document_embeddings.dim()}."
+                        )
+                    if query_embeddings.size(0) not in [1, document_embeddings.size(0)]:
+                        raise ValueError(
+                            "There should be either one query or queries equal to the number of documents."
+                        )
+
+                    # Transpose the query embeddings to align for matrix multiplication
+                    transposed_query_embeddings = query_embeddings.permute(0, 2, 1)
+                    # Compute similarity scores using batch matrix multiplication
+                    computed_scores = torch.matmul(
+                        document_embeddings, transposed_query_embeddings
+                    )
+                    # Apply max pooling to extract the highest semantic similarity across each document's sequence
+                    maximum_scores = torch.max(computed_scores, dim=1).values
+
+                    # Sum up the maximum scores across features to get the overall document relevance scores
+                    final_scores = maximum_scores.sum(dim=1)
+
+                    normalized_scores = torch.softmax(final_scores, dim=0)
+
+                    return normalized_scores.detach().cpu().numpy().astype(np.float32)
+
+                def predict(self, sentences):
+
+                    query = sentences[0][0]
+                    docs = [i[1] for i in sentences]
+
+                    # Embedding the documents
+                    embedded_docs = self.ckpt.docFromText(docs, bsize=32)[0]
+                    # Embedding the queries
+                    embedded_queries = self.ckpt.queryFromText([query], bsize=32)
+                    embedded_query = embedded_queries[0]
+
+                    # Calculate retrieval scores for the query against all documents
+                    scores = self.calculate_similarity_scores(
+                        embedded_query.unsqueeze(0), embedded_docs
+                    )
+
+                    return scores
+
+            try:
+                app.state.sentence_transformer_rf = ColBERT(
+                    get_model_path(reranking_model, auto_update)
+                )
+            except Exception as e:
+                log.error(f"ColBERT: {e}")
+                app.state.sentence_transformer_rf = None
+                app.state.config.ENABLE_RAG_HYBRID_SEARCH = False
+        else:
+            import sentence_transformers
+
+            try:
+                app.state.sentence_transformer_rf = sentence_transformers.CrossEncoder(
+                    get_model_path(reranking_model, auto_update),
+                    device=DEVICE_TYPE,
+                    trust_remote_code=RAG_RERANKING_MODEL_TRUST_REMOTE_CODE,
+                )
+            except:
+                log.error("CrossEncoder error")
+                app.state.sentence_transformer_rf = None
+                app.state.config.ENABLE_RAG_HYBRID_SEARCH = False
     else:
         app.state.sentence_transformer_rf = None
 
@@ -240,12 +322,9 @@ app.state.EMBEDDING_FUNCTION = get_embedding_function(
     app.state.config.RAG_EMBEDDING_OPENAI_BATCH_SIZE,
 )
 
-origins = ["*"]
-
-
 app.add_middleware(
     CORSMiddleware,
-    allow_origins=origins,
+    allow_origins=CORS_ALLOW_ORIGIN,
     allow_credentials=True,
     allow_methods=["*"],
     allow_headers=["*"],
@@ -376,7 +455,7 @@ async def update_reranking_config(
     try:
         app.state.config.RAG_RERANKING_MODEL = form_data.reranking_model
 
-        update_reranking_model(app.state.config.RAG_RERANKING_MODEL), True
+        update_reranking_model(app.state.config.RAG_RERANKING_MODEL, True)
 
         return {
             "status": True,
@@ -395,6 +474,10 @@ async def get_rag_config(user=Depends(get_admin_user)):
     return {
         "status": True,
         "pdf_extract_images": app.state.config.PDF_EXTRACT_IMAGES,
+        "file": {
+            "max_size": app.state.config.FILE_MAX_SIZE,
+            "max_count": app.state.config.FILE_MAX_COUNT,
+        },
         "content_extraction": {
             "engine": app.state.config.CONTENT_EXTRACTION_ENGINE,
             "tika_server_url": app.state.config.TIKA_SERVER_URL,
@@ -421,6 +504,8 @@ async def get_rag_config(user=Depends(get_admin_user)):
                 "serper_api_key": app.state.config.SERPER_API_KEY,
                 "serply_api_key": app.state.config.SERPLY_API_KEY,
                 "tavily_api_key": app.state.config.TAVILY_API_KEY,
+                "searchapi_api_key": app.state.config.SEARCHAPI_API_KEY,
+                "seaarchapi_engine": app.state.config.SEARCHAPI_ENGINE,
                 "result_count": app.state.config.RAG_WEB_SEARCH_RESULT_COUNT,
                 "concurrent_requests": app.state.config.RAG_WEB_SEARCH_CONCURRENT_REQUESTS,
             },
@@ -428,6 +513,11 @@ async def get_rag_config(user=Depends(get_admin_user)):
     }
 
 
+class FileConfig(BaseModel):
+    max_size: Optional[int] = None
+    max_count: Optional[int] = None
+
+
 class ContentExtractionConfig(BaseModel):
     engine: str = ""
     tika_server_url: Optional[str] = None
@@ -439,7 +529,7 @@ class ChunkParamUpdateForm(BaseModel):
 
 
 class YoutubeLoaderConfig(BaseModel):
-    language: List[str]
+    language: list[str]
     translation: Optional[str] = None
 
 
@@ -455,6 +545,8 @@ class WebSearchConfig(BaseModel):
     serper_api_key: Optional[str] = None
     serply_api_key: Optional[str] = None
     tavily_api_key: Optional[str] = None
+    searchapi_api_key: Optional[str] = None
+    searchapi_engine: Optional[str] = None
     result_count: Optional[int] = None
     concurrent_requests: Optional[int] = None
 
@@ -466,6 +558,7 @@ class WebConfig(BaseModel):
 
 class ConfigUpdateForm(BaseModel):
     pdf_extract_images: Optional[bool] = None
+    file: Optional[FileConfig] = None
     content_extraction: Optional[ContentExtractionConfig] = None
     chunk: Optional[ChunkParamUpdateForm] = None
     youtube: Optional[YoutubeLoaderConfig] = None
@@ -480,6 +573,10 @@ async def update_rag_config(form_data: ConfigUpdateForm, user=Depends(get_admin_
         else app.state.config.PDF_EXTRACT_IMAGES
     )
 
+    if form_data.file is not None:
+        app.state.config.FILE_MAX_SIZE = form_data.file.max_size
+        app.state.config.FILE_MAX_COUNT = form_data.file.max_count
+
     if form_data.content_extraction is not None:
         log.info(f"Updating text settings: {form_data.content_extraction}")
         app.state.config.CONTENT_EXTRACTION_ENGINE = form_data.content_extraction.engine
@@ -513,6 +610,8 @@ async def update_rag_config(form_data: ConfigUpdateForm, user=Depends(get_admin_
         app.state.config.SERPER_API_KEY = form_data.web.search.serper_api_key
         app.state.config.SERPLY_API_KEY = form_data.web.search.serply_api_key
         app.state.config.TAVILY_API_KEY = form_data.web.search.tavily_api_key
+        app.state.config.SEARCHAPI_API_KEY = form_data.web.search.searchapi_api_key
+        app.state.config.SEARCHAPI_ENGINE = form_data.web.search.searchapi_engine
         app.state.config.RAG_WEB_SEARCH_RESULT_COUNT = form_data.web.search.result_count
         app.state.config.RAG_WEB_SEARCH_CONCURRENT_REQUESTS = (
             form_data.web.search.concurrent_requests
@@ -521,6 +620,10 @@ async def update_rag_config(form_data: ConfigUpdateForm, user=Depends(get_admin_
     return {
         "status": True,
         "pdf_extract_images": app.state.config.PDF_EXTRACT_IMAGES,
+        "file": {
+            "max_size": app.state.config.FILE_MAX_SIZE,
+            "max_count": app.state.config.FILE_MAX_COUNT,
+        },
         "content_extraction": {
             "engine": app.state.config.CONTENT_EXTRACTION_ENGINE,
             "tika_server_url": app.state.config.TIKA_SERVER_URL,
@@ -546,6 +649,8 @@ async def update_rag_config(form_data: ConfigUpdateForm, user=Depends(get_admin_
                 "serpstack_https": app.state.config.SERPSTACK_HTTPS,
                 "serper_api_key": app.state.config.SERPER_API_KEY,
                 "serply_api_key": app.state.config.SERPLY_API_KEY,
+                "serachapi_api_key": app.state.config.SEARCHAPI_API_KEY,
+                "searchapi_engine": app.state.config.SEARCHAPI_ENGINE,
                 "tavily_api_key": app.state.config.TAVILY_API_KEY,
                 "result_count": app.state.config.RAG_WEB_SEARCH_RESULT_COUNT,
                 "concurrent_requests": app.state.config.RAG_WEB_SEARCH_CONCURRENT_REQUESTS,
@@ -585,13 +690,14 @@ async def update_query_settings(
     form_data: QuerySettingsForm, user=Depends(get_admin_user)
 ):
     app.state.config.RAG_TEMPLATE = (
-        form_data.template if form_data.template else RAG_TEMPLATE
+        form_data.template if form_data.template != "" else DEFAULT_RAG_TEMPLATE
     )
     app.state.config.TOP_K = form_data.k if form_data.k else 4
     app.state.config.RELEVANCE_THRESHOLD = form_data.r if form_data.r else 0.0
     app.state.config.ENABLE_RAG_HYBRID_SEARCH = (
         form_data.hybrid if form_data.hybrid else False
     )
+
     return {
         "status": True,
         "template": app.state.config.RAG_TEMPLATE,
@@ -642,7 +748,7 @@ def query_doc_handler(
 
 
 class QueryCollectionsForm(BaseModel):
-    collection_names: List[str]
+    collection_names: list[str]
     query: str
     k: Optional[int] = None
     r: Optional[float] = None
@@ -796,6 +902,7 @@ def search_web(engine: str, query: str) -> list[SearchResult]:
     - SERPER_API_KEY
     - SERPLY_API_KEY
     - TAVILY_API_KEY
+    - SEARCHAPI_API_KEY + SEARCHAPI_ENGINE (by default `google`)
     Args:
         query (str): The query to search for
     """
@@ -883,6 +990,17 @@ def search_web(engine: str, query: str) -> list[SearchResult]:
             )
         else:
             raise Exception("No TAVILY_API_KEY found in environment variables")
+    elif engine == "searchapi":
+        if app.state.config.SEARCHAPI_API_KEY:
+            return search_searchapi(
+                app.state.config.SEARCHAPI_API_KEY,
+                app.state.config.SEARCHAPI_ENGINE,
+                query,
+                app.state.config.RAG_WEB_SEARCH_RESULT_COUNT,
+                app.state.config.RAG_WEB_SEARCH_DOMAIN_FILTER_LIST,
+            )
+        else:
+            raise Exception("No SEARCHAPI_API_KEY found in environment variables")
     elif engine == "jina":
         return search_jina(query, app.state.config.RAG_WEB_SEARCH_RESULT_COUNT)
     else:
@@ -933,7 +1051,6 @@ def store_web_search(form_data: SearchForm, user=Depends(get_verified_user)):
 def store_data_in_vector_db(
     data, collection_name, metadata: Optional[dict] = None, overwrite: bool = False
 ) -> bool:
-
     text_splitter = RecursiveCharacterTextSplitter(
         chunk_size=app.state.config.CHUNK_SIZE,
         chunk_overlap=app.state.config.CHUNK_OVERLAP,
@@ -978,41 +1095,39 @@ def store_docs_in_vector_db(
 
     try:
         if overwrite:
-            for collection in CHROMA_CLIENT.list_collections():
-                if collection_name == collection.name:
-                    log.info(f"deleting existing collection {collection_name}")
-                    CHROMA_CLIENT.delete_collection(name=collection_name)
+            if VECTOR_DB_CLIENT.has_collection(collection_name=collection_name):
+                log.info(f"deleting existing collection {collection_name}")
+                VECTOR_DB_CLIENT.delete_collection(collection_name=collection_name)
 
-        collection = CHROMA_CLIENT.create_collection(name=collection_name)
-
-        embedding_func = get_embedding_function(
-            app.state.config.RAG_EMBEDDING_ENGINE,
-            app.state.config.RAG_EMBEDDING_MODEL,
-            app.state.sentence_transformer_ef,
-            app.state.config.OPENAI_API_KEY,
-            app.state.config.OPENAI_API_BASE_URL,
-            app.state.config.RAG_EMBEDDING_OPENAI_BATCH_SIZE,
-        )
-
-        embedding_texts = list(map(lambda x: x.replace("\n", " "), texts))
-        embeddings = embedding_func(embedding_texts)
-
-        for batch in create_batches(
-            api=CHROMA_CLIENT,
-            ids=[str(uuid.uuid4()) for _ in texts],
-            metadatas=metadatas,
-            embeddings=embeddings,
-            documents=texts,
-        ):
-            collection.add(*batch)
-
-        return True
-    except Exception as e:
-        if e.__class__.__name__ == "UniqueConstraintError":
+        if VECTOR_DB_CLIENT.has_collection(collection_name=collection_name):
+            log.info(f"collection {collection_name} already exists")
             return True
+        else:
+            embedding_function = get_embedding_function(
+                app.state.config.RAG_EMBEDDING_ENGINE,
+                app.state.config.RAG_EMBEDDING_MODEL,
+                app.state.sentence_transformer_ef,
+                app.state.config.OPENAI_API_KEY,
+                app.state.config.OPENAI_API_BASE_URL,
+                app.state.config.RAG_EMBEDDING_OPENAI_BATCH_SIZE,
+            )
 
+            VECTOR_DB_CLIENT.insert(
+                collection_name=collection_name,
+                items=[
+                    {
+                        "id": str(uuid.uuid4()),
+                        "text": text,
+                        "vector": embedding_function(text.replace("\n", " ")),
+                        "metadata": metadatas[idx],
+                    }
+                    for idx, text in enumerate(texts)
+                ],
+            )
+
+            return True
+    except Exception as e:
         log.exception(e)
-
         return False
 
 
@@ -1021,7 +1136,7 @@ class TikaLoader:
         self.file_path = file_path
         self.mime_type = mime_type
 
-    def load(self) -> List[Document]:
+    def load(self) -> list[Document]:
         with open(self.file_path, "rb") as f:
             data = f.read()
 
@@ -1138,7 +1253,7 @@ def get_loader(filename: str, file_content_type: str, file_path: str):
         elif (
             file_content_type
             == "application/vnd.openxmlformats-officedocument.wordprocessingml.document"
-            or file_ext in ["doc", "docx"]
+            or file_ext == "docx"
         ):
             loader = Docx2txtLoader(file_path)
         elif file_content_type in [
@@ -1185,7 +1300,7 @@ def store_doc(
             f.close()
 
         f = open(file_path, "rb")
-        if collection_name == None:
+        if collection_name is None:
             collection_name = calculate_sha256(f)[:63]
         f.close()
 
@@ -1238,7 +1353,7 @@ def process_doc(
         f = open(file_path, "rb")
 
         collection_name = form_data.collection_name
-        if collection_name == None:
+        if collection_name is None:
             collection_name = calculate_sha256(f)[:63]
         f.close()
 
@@ -1294,9 +1409,8 @@ def store_text(
     form_data: TextRAGForm,
     user=Depends(get_verified_user),
 ):
-
     collection_name = form_data.collection_name
-    if collection_name == None:
+    if collection_name is None:
         collection_name = calculate_sha256_string(form_data.content)
 
     result = store_text_in_vector_db(
@@ -1339,7 +1453,7 @@ def scan_docs_dir(user=Depends(get_admin_user)):
                         sanitized_filename = sanitize_filename(filename)
                         doc = Documents.get_doc_by_name(sanitized_filename)
 
-                        if doc == None:
+                        if doc is None:
                             doc = Documents.insert_new_doc(
                                 user.id,
                                 DocumentForm(
@@ -1375,12 +1489,12 @@ def scan_docs_dir(user=Depends(get_admin_user)):
     return True
 
 
-@app.get("/reset/db")
+@app.post("/reset/db")
 def reset_vector_db(user=Depends(get_admin_user)):
-    CHROMA_CLIENT.reset()
+    VECTOR_DB_CLIENT.reset()
 
 
-@app.get("/reset/uploads")
+@app.post("/reset/uploads")
 def reset_upload_dir(user=Depends(get_admin_user)) -> bool:
     folder = f"{UPLOAD_DIR}"
     try:
@@ -1404,7 +1518,7 @@ def reset_upload_dir(user=Depends(get_admin_user)) -> bool:
     return True
 
 
-@app.get("/reset")
+@app.post("/reset")
 def reset(user=Depends(get_admin_user)) -> bool:
     folder = f"{UPLOAD_DIR}"
     for filename in os.listdir(folder):
@@ -1418,7 +1532,7 @@ def reset(user=Depends(get_admin_user)) -> bool:
             log.error("Failed to delete %s. Reason: %s" % (file_path, e))
 
     try:
-        CHROMA_CLIENT.reset()
+        VECTOR_DB_CLIENT.reset()
     except Exception as e:
         log.exception(e)
 

backend/open_webui/apps/rag/search/brave.py

@@ -1,16 +1,16 @@
 import logging
-from typing import List, Optional
-import requests
+from typing import Optional
 
-from apps.rag.search.main import SearchResult, get_filtered_results
-from config import SRC_LOG_LEVELS
+import requests
+from open_webui.apps.rag.search.main import SearchResult, get_filtered_results
+from open_webui.env import SRC_LOG_LEVELS
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["RAG"])
 
 
 def search_brave(
-    api_key: str, query: str, count: int, filter_list: Optional[List[str]] = None
+    api_key: str, query: str, count: int, filter_list: Optional[list[str]] = None
 ) -> list[SearchResult]:
     """Search using Brave's Search API and return the results as a list of SearchResult objects.
 

backend/open_webui/apps/rag/search/duckduckgo.py

@@ -1,15 +1,16 @@
 import logging
-from typing import List, Optional
-from apps.rag.search.main import SearchResult, get_filtered_results
+from typing import Optional
+
+from open_webui.apps.rag.search.main import SearchResult, get_filtered_results
 from duckduckgo_search import DDGS
-from config import SRC_LOG_LEVELS
+from open_webui.env import SRC_LOG_LEVELS
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["RAG"])
 
 
 def search_duckduckgo(
-    query: str, count: int, filter_list: Optional[List[str]] = None
+    query: str, count: int, filter_list: Optional[list[str]] = None
 ) -> list[SearchResult]:
     """
     Search using DuckDuckGo's Search API and return the results as a list of SearchResult objects.
@@ -18,7 +19,7 @@ def search_duckduckgo(
         count (int): The number of results to return
 
     Returns:
-        List[SearchResult]: A list of search results
+        list[SearchResult]: A list of search results
     """
     # Use the DDGS context manager to create a DDGS object
     with DDGS() as ddgs:

backend/open_webui/apps/rag/search/google_pse.py

@@ -1,10 +1,9 @@
-import json
 import logging
-from typing import List, Optional
-import requests
+from typing import Optional
 
-from apps.rag.search.main import SearchResult, get_filtered_results
-from config import SRC_LOG_LEVELS
+import requests
+from open_webui.apps.rag.search.main import SearchResult, get_filtered_results
+from open_webui.env import SRC_LOG_LEVELS
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["RAG"])
@@ -15,7 +14,7 @@ def search_google_pse(
     search_engine_id: str,
     query: str,
     count: int,
-    filter_list: Optional[List[str]] = None,
+    filter_list: Optional[list[str]] = None,
 ) -> list[SearchResult]:
     """Search using Google's Programmable Search Engine API and return the results as a list of SearchResult objects.
 

backend/open_webui/apps/rag/search/jina_search.py

@@ -1,9 +1,9 @@
 import logging
-import requests
-from yarl import URL
 
-from apps.rag.search.main import SearchResult
-from config import SRC_LOG_LEVELS
+import requests
+from open_webui.apps.rag.search.main import SearchResult
+from open_webui.env import SRC_LOG_LEVELS
+from yarl import URL
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["RAG"])
@@ -17,7 +17,7 @@ def search_jina(query: str, count: int) -> list[SearchResult]:
         count (int): The number of results to return
 
     Returns:
-        List[SearchResult]: A list of search results
+        list[SearchResult]: A list of search results
     """
     jina_search_endpoint = "https://s.jina.ai/"
     headers = {

backend/open_webui/apps/rag/search/main.py

@@ -1,5 +1,6 @@
 from typing import Optional
 from urllib.parse import urlparse
+
 from pydantic import BaseModel
 
 
@@ -8,7 +9,8 @@ def get_filtered_results(results, filter_list):
         return results
     filtered_results = []
     for result in results:
-        domain = urlparse(result["url"]).netloc
+        url = result.get("url") or result.get("link", "")
+        domain = urlparse(url).netloc
         if any(domain.endswith(filtered_domain) for filtered_domain in filter_list):
             filtered_results.append(result)
     return filtered_results

backend/open_webui/apps/rag/search/searchapi.py

@@ -0,0 +1,48 @@
+import logging
+from typing import Optional
+from urllib.parse import urlencode
+
+import requests
+from open_webui.apps.rag.search.main import SearchResult, get_filtered_results
+from open_webui.env import SRC_LOG_LEVELS
+
+log = logging.getLogger(__name__)
+log.setLevel(SRC_LOG_LEVELS["RAG"])
+
+
+def search_searchapi(
+    api_key: str,
+    engine: str,
+    query: str,
+    count: int,
+    filter_list: Optional[list[str]] = None,
+) -> list[SearchResult]:
+    """Search using searchapi.io's API and return the results as a list of SearchResult objects.
+
+    Args:
+      api_key (str): A searchapi.io API key
+      query (str): The query to search for
+    """
+    url = "https://www.searchapi.io/api/v1/search"
+
+    engine = engine or "google"
+
+    payload = {"engine": engine, "q": query, "api_key": api_key}
+
+    url = f"{url}?{urlencode(payload)}"
+    response = requests.request("GET", url)
+
+    json_response = response.json()
+    log.info(f"results from searchapi search: {json_response}")
+
+    results = sorted(
+        json_response.get("organic_results", []), key=lambda x: x.get("position", 0)
+    )
+    if filter_list:
+        results = get_filtered_results(results, filter_list)
+    return [
+        SearchResult(
+            link=result["link"], title=result["title"], snippet=result["snippet"]
+        )
+        for result in results[:count]
+    ]

backend/open_webui/apps/rag/search/searxng.py

@@ -1,10 +1,9 @@
 import logging
+from typing import Optional
+
 import requests
-
-from typing import List, Optional
-
-from apps.rag.search.main import SearchResult, get_filtered_results
-from config import SRC_LOG_LEVELS
+from open_webui.apps.rag.search.main import SearchResult, get_filtered_results
+from open_webui.env import SRC_LOG_LEVELS
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["RAG"])
@@ -14,9 +13,9 @@ def search_searxng(
     query_url: str,
     query: str,
     count: int,
-    filter_list: Optional[List[str]] = None,
+    filter_list: Optional[list[str]] = None,
     **kwargs,
-) -> List[SearchResult]:
+) -> list[SearchResult]:
     """
     Search a SearXNG instance for a given query and return the results as a list of SearchResult objects.
 
@@ -31,10 +30,10 @@ def search_searxng(
         language (str): Language filter for the search results; e.g., "en-US". Defaults to an empty string.
         safesearch (int): Safe search filter for safer web results; 0 = off, 1 = moderate, 2 = strict. Defaults to 1 (moderate).
         time_range (str): Time range for filtering results by date; e.g., "2023-04-05..today" or "all-time". Defaults to ''.
-        categories: (Optional[List[str]]): Specific categories within which the search should be performed, defaulting to an empty string if not provided.
+        categories: (Optional[list[str]]): Specific categories within which the search should be performed, defaulting to an empty string if not provided.
 
     Returns:
-        List[SearchResult]: A list of SearchResults sorted by relevance score in descending order.
+        list[SearchResult]: A list of SearchResults sorted by relevance score in descending order.
 
     Raise:
         requests.exceptions.RequestException: If a request error occurs during the search process.

backend/open_webui/apps/rag/search/serper.py

@@ -1,17 +1,17 @@
 import json
 import logging
-from typing import List, Optional
-import requests
+from typing import Optional
 
-from apps.rag.search.main import SearchResult, get_filtered_results
-from config import SRC_LOG_LEVELS
+import requests
+from open_webui.apps.rag.search.main import SearchResult, get_filtered_results
+from open_webui.env import SRC_LOG_LEVELS
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["RAG"])
 
 
 def search_serper(
-    api_key: str, query: str, count: int, filter_list: Optional[List[str]] = None
+    api_key: str, query: str, count: int, filter_list: Optional[list[str]] = None
 ) -> list[SearchResult]:
     """Search using serper.dev's API and return the results as a list of SearchResult objects.
 

backend/open_webui/apps/rag/search/serply.py

@@ -1,11 +1,10 @@
-import json
 import logging
-from typing import List, Optional
-import requests
+from typing import Optional
 from urllib.parse import urlencode
 
-from apps.rag.search.main import SearchResult, get_filtered_results
-from config import SRC_LOG_LEVELS
+import requests
+from open_webui.apps.rag.search.main import SearchResult, get_filtered_results
+from open_webui.env import SRC_LOG_LEVELS
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["RAG"])
@@ -19,7 +18,7 @@ def search_serply(
     limit: int = 10,
     device_type: str = "desktop",
     proxy_location: str = "US",
-    filter_list: Optional[List[str]] = None,
+    filter_list: Optional[list[str]] = None,
 ) -> list[SearchResult]:
     """Search using serper.dev's API and return the results as a list of SearchResult objects.
 

backend/open_webui/apps/rag/search/serpstack.py

@@ -1,10 +1,9 @@
-import json
 import logging
-from typing import List, Optional
-import requests
+from typing import Optional
 
-from apps.rag.search.main import SearchResult, get_filtered_results
-from config import SRC_LOG_LEVELS
+import requests
+from open_webui.apps.rag.search.main import SearchResult, get_filtered_results
+from open_webui.env import SRC_LOG_LEVELS
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["RAG"])
@@ -14,7 +13,7 @@ def search_serpstack(
     api_key: str,
     query: str,
     count: int,
-    filter_list: Optional[List[str]] = None,
+    filter_list: Optional[list[str]] = None,
     https_enabled: bool = True,
 ) -> list[SearchResult]:
     """Search using serpstack.com's and return the results as a list of SearchResult objects.

backend/open_webui/apps/rag/search/tavily.py

@@ -1,9 +1,8 @@
 import logging
 
 import requests
-
-from apps.rag.search.main import SearchResult
-from config import SRC_LOG_LEVELS
+from open_webui.apps.rag.search.main import SearchResult
+from open_webui.env import SRC_LOG_LEVELS
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["RAG"])
@@ -17,7 +16,7 @@ def search_tavily(api_key: str, query: str, count: int) -> list[SearchResult]:
         query (str): The query to search for
 
     Returns:
-        List[SearchResult]: A list of search results
+        list[SearchResult]: A list of search results
     """
     url = "https://api.tavily.com/search"
     data = {"query": query, "api_key": api_key}

backend/open_webui/apps/rag/utils.py

@@ -1,32 +1,68 @@
-import os
 import logging
+import os
+import uuid
+from typing import Optional, Union
+
 import requests
 
-from typing import List, Union
-
-from apps.ollama.main import (
-    generate_ollama_embeddings,
-    GenerateEmbeddingsForm,
-)
-
 from huggingface_hub import snapshot_download
-
-from langchain_core.documents import Document
+from langchain.retrievers import ContextualCompressionRetriever, EnsembleRetriever
 from langchain_community.retrievers import BM25Retriever
-from langchain.retrievers import (
-    ContextualCompressionRetriever,
-    EnsembleRetriever,
+from langchain_core.documents import Document
+
+
+from open_webui.apps.ollama.main import (
+    GenerateEmbeddingsForm,
+    generate_ollama_embeddings,
 )
+from open_webui.apps.rag.vector.connector import VECTOR_DB_CLIENT
+from open_webui.utils.misc import get_last_user_message
 
-from typing import Optional
+from open_webui.env import SRC_LOG_LEVELS
 
-from utils.misc import get_last_user_message, add_or_update_system_message
-from config import SRC_LOG_LEVELS, CHROMA_CLIENT
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["RAG"])
 
 
+from typing import Any
+
+from langchain_core.callbacks import CallbackManagerForRetrieverRun
+from langchain_core.retrievers import BaseRetriever
+
+
+class VectorSearchRetriever(BaseRetriever):
+    collection_name: Any
+    embedding_function: Any
+    top_k: int
+
+    def _get_relevant_documents(
+        self,
+        query: str,
+        *,
+        run_manager: CallbackManagerForRetrieverRun,
+    ) -> list[Document]:
+        result = VECTOR_DB_CLIENT.search(
+            collection_name=self.collection_name,
+            vectors=[self.embedding_function(query)],
+            limit=self.top_k,
+        )
+
+        ids = result.ids[0]
+        metadatas = result.metadatas[0]
+        documents = result.documents[0]
+
+        results = []
+        for idx in range(len(ids)):
+            results.append(
+                Document(
+                    metadata=metadatas[idx],
+                    page_content=documents[idx],
+                )
+            )
+        return results
+
+
 def query_doc(
     collection_name: str,
     query: str,
@@ -34,17 +70,18 @@ def query_doc(
     k: int,
 ):
     try:
-        collection = CHROMA_CLIENT.get_collection(name=collection_name)
-        query_embeddings = embedding_function(query)
-
-        result = collection.query(
-            query_embeddings=[query_embeddings],
-            n_results=k,
+        result = VECTOR_DB_CLIENT.search(
+            collection_name=collection_name,
+            vectors=[embedding_function(query)],
+            limit=k,
         )
 
+        print("result", result)
+
         log.info(f"query_doc:result {result}")
         return result
     except Exception as e:
+        print(e)
         raise e
 
 
@@ -55,27 +92,25 @@ def query_doc_with_hybrid_search(
     k: int,
     reranking_function,
     r: float,
-):
+) -> dict:
     try:
-        collection = CHROMA_CLIENT.get_collection(name=collection_name)
-        documents = collection.get()  # get all documents
+        result = VECTOR_DB_CLIENT.get(collection_name=collection_name)
 
         bm25_retriever = BM25Retriever.from_texts(
-            texts=documents.get("documents"),
-            metadatas=documents.get("metadatas"),
+            texts=result.documents[0],
+            metadatas=result.metadatas[0],
         )
         bm25_retriever.k = k
 
-        chroma_retriever = ChromaRetriever(
-            collection=collection,
+        vector_search_retriever = VectorSearchRetriever(
+            collection_name=collection_name,
             embedding_function=embedding_function,
-            top_n=k,
+            top_k=k,
         )
 
         ensemble_retriever = EnsembleRetriever(
-            retrievers=[bm25_retriever, chroma_retriever], weights=[0.5, 0.5]
+            retrievers=[bm25_retriever, vector_search_retriever], weights=[0.5, 0.5]
         )
-
         compressor = RerankCompressor(
             embedding_function=embedding_function,
             top_n=k,
@@ -100,7 +135,9 @@ def query_doc_with_hybrid_search(
         raise e
 
 
-def merge_and_sort_query_results(query_results, k, reverse=False):
+def merge_and_sort_query_results(
+    query_results: list[dict], k: int, reverse: bool = False
+) -> list[dict]:
     # Initialize lists to store combined data
     combined_distances = []
     combined_documents = []
@@ -142,35 +179,40 @@ def merge_and_sort_query_results(query_results, k, reverse=False):
 
 
 def query_collection(
-    collection_names: List[str],
+    collection_names: list[str],
     query: str,
     embedding_function,
     k: int,
-):
+) -> dict:
     results = []
     for collection_name in collection_names:
-        try:
-            result = query_doc(
-                collection_name=collection_name,
-                query=query,
-                k=k,
-                embedding_function=embedding_function,
-            )
-            results.append(result)
-        except:
+        if collection_name:
+            try:
+                result = query_doc(
+                    collection_name=collection_name,
+                    query=query,
+                    k=k,
+                    embedding_function=embedding_function,
+                )
+                results.append(result.model_dump())
+            except Exception as e:
+                log.exception(f"Error when querying the collection: {e}")
+        else:
             pass
+
     return merge_and_sort_query_results(results, k=k)
 
 
 def query_collection_with_hybrid_search(
-    collection_names: List[str],
+    collection_names: list[str],
     query: str,
     embedding_function,
     k: int,
     reranking_function,
     r: float,
-):
+) -> dict:
     results = []
+    error = False
     for collection_name in collection_names:
         try:
             result = query_doc_with_hybrid_search(
@@ -182,14 +224,39 @@ def query_collection_with_hybrid_search(
                 r=r,
             )
             results.append(result)
-        except:
-            pass
+        except Exception as e:
+            log.exception(
+                "Error when querying the collection with " f"hybrid_search: {e}"
+            )
+            error = True
+
+    if error:
+        raise Exception(
+            "Hybrid search failed for all collections. Using Non hybrid search as fallback."
+        )
+
     return merge_and_sort_query_results(results, k=k, reverse=True)
 
 
 def rag_template(template: str, context: str, query: str):
-    template = template.replace("[context]", context)
-    template = template.replace("[query]", query)
+    count = template.count("[context]")
+    assert "[context]" in template, "RAG template does not contain '[context]'"
+
+    if "<context>" in context and "</context>" in context:
+        log.debug(
+            "WARNING: Potential prompt injection attack: the RAG "
+            "context contains '<context>' and '</context>'. This might be "
+            "nothing, or the user might be trying to hack something."
+        )
+
+    if "[query]" in context:
+        query_placeholder = f"[query-{str(uuid.uuid4())}]"
+        template = template.replace("[query]", query_placeholder)
+        template = template.replace("[context]", context)
+        template = template.replace(query_placeholder, query)
+    else:
+        template = template.replace("[context]", context)
+        template = template.replace("[query]", query)
     return template
 
 
@@ -257,7 +324,7 @@ def get_rag_context(
         collection_names = (
             file["collection_names"]
             if file["type"] == "collection"
-            else [file["collection_name"]]
+            else [file["collection_name"]] if file["collection_name"] else []
         )
 
         collection_names = set(collection_names).difference(extracted_collections)
@@ -266,19 +333,27 @@ def get_rag_context(
             continue
 
         try:
+            context = None
             if file["type"] == "text":
                 context = file["content"]
             else:
                 if hybrid_search:
-                    context = query_collection_with_hybrid_search(
-                        collection_names=collection_names,
-                        query=query,
-                        embedding_function=embedding_function,
-                        k=k,
-                        reranking_function=reranking_function,
-                        r=r,
-                    )
-                else:
+                    try:
+                        context = query_collection_with_hybrid_search(
+                            collection_names=collection_names,
+                            query=query,
+                            embedding_function=embedding_function,
+                            k=k,
+                            reranking_function=reranking_function,
+                            r=r,
+                        )
+                    except Exception as e:
+                        log.debug(
+                            "Error when using hybrid search, using"
+                            " non hybrid search as fallback."
+                        )
+
+                if (not hybrid_search) or (context is None):
                     context = query_collection(
                         collection_names=collection_names,
                         query=query,
@@ -287,7 +362,6 @@ def get_rag_context(
                     )
         except Exception as e:
             log.exception(e)
-            context = None
 
         if context:
             relevant_contexts.append({**context, "source": file})
@@ -395,52 +469,11 @@ def generate_openai_batch_embeddings(
         return None
 
 
-from typing import Any
-
-from langchain_core.retrievers import BaseRetriever
-from langchain_core.callbacks import CallbackManagerForRetrieverRun
-
-
-class ChromaRetriever(BaseRetriever):
-    collection: Any
-    embedding_function: Any
-    top_n: int
-
-    def _get_relevant_documents(
-        self,
-        query: str,
-        *,
-        run_manager: CallbackManagerForRetrieverRun,
-    ) -> List[Document]:
-        query_embeddings = self.embedding_function(query)
-
-        results = self.collection.query(
-            query_embeddings=[query_embeddings],
-            n_results=self.top_n,
-        )
-
-        ids = results["ids"][0]
-        metadatas = results["metadatas"][0]
-        documents = results["documents"][0]
-
-        results = []
-        for idx in range(len(ids)):
-            results.append(
-                Document(
-                    metadata=metadatas[idx],
-                    page_content=documents[idx],
-                )
-            )
-        return results
-
-
 import operator
-
 from typing import Optional, Sequence
 
-from langchain_core.documents import BaseDocumentCompressor, Document
 from langchain_core.callbacks import Callbacks
-from langchain_core.pydantic_v1 import Extra
+from langchain_core.documents import BaseDocumentCompressor, Document
 
 
 class RerankCompressor(BaseDocumentCompressor):
@@ -450,7 +483,7 @@ class RerankCompressor(BaseDocumentCompressor):
     r_score: float
 
     class Config:
-        extra = Extra.forbid
+        extra = "forbid"
         arbitrary_types_allowed = True
 
     def compress_documents(

backend/open_webui/apps/rag/vector/connector.py

@@ -0,0 +1,10 @@
+from open_webui.apps.rag.vector.dbs.chroma import ChromaClient
+from open_webui.apps.rag.vector.dbs.milvus import MilvusClient
+
+
+from open_webui.config import VECTOR_DB
+
+if VECTOR_DB == "milvus":
+    VECTOR_DB_CLIENT = MilvusClient()
+else:
+    VECTOR_DB_CLIENT = ChromaClient()

backend/open_webui/apps/rag/vector/dbs/chroma.py

@@ -0,0 +1,122 @@
+import chromadb
+from chromadb import Settings
+from chromadb.utils.batch_utils import create_batches
+
+from typing import Optional
+
+from open_webui.apps.rag.vector.main import VectorItem, SearchResult, GetResult
+from open_webui.config import (
+    CHROMA_DATA_PATH,
+    CHROMA_HTTP_HOST,
+    CHROMA_HTTP_PORT,
+    CHROMA_HTTP_HEADERS,
+    CHROMA_HTTP_SSL,
+    CHROMA_TENANT,
+    CHROMA_DATABASE,
+)
+
+
+class ChromaClient:
+    def __init__(self):
+        if CHROMA_HTTP_HOST != "":
+            self.client = chromadb.HttpClient(
+                host=CHROMA_HTTP_HOST,
+                port=CHROMA_HTTP_PORT,
+                headers=CHROMA_HTTP_HEADERS,
+                ssl=CHROMA_HTTP_SSL,
+                tenant=CHROMA_TENANT,
+                database=CHROMA_DATABASE,
+                settings=Settings(allow_reset=True, anonymized_telemetry=False),
+            )
+        else:
+            self.client = chromadb.PersistentClient(
+                path=CHROMA_DATA_PATH,
+                settings=Settings(allow_reset=True, anonymized_telemetry=False),
+                tenant=CHROMA_TENANT,
+                database=CHROMA_DATABASE,
+            )
+
+    def has_collection(self, collection_name: str) -> bool:
+        # Check if the collection exists based on the collection name.
+        collections = self.client.list_collections()
+        return collection_name in [collection.name for collection in collections]
+
+    def delete_collection(self, collection_name: str):
+        # Delete the collection based on the collection name.
+        return self.client.delete_collection(name=collection_name)
+
+    def search(
+        self, collection_name: str, vectors: list[list[float | int]], limit: int
+    ) -> Optional[SearchResult]:
+        # Search for the nearest neighbor items based on the vectors and return 'limit' number of results.
+        collection = self.client.get_collection(name=collection_name)
+        if collection:
+            result = collection.query(
+                query_embeddings=vectors,
+                n_results=limit,
+            )
+
+            return SearchResult(
+                **{
+                    "ids": result["ids"],
+                    "distances": result["distances"],
+                    "documents": result["documents"],
+                    "metadatas": result["metadatas"],
+                }
+            )
+        return None
+
+    def get(self, collection_name: str) -> Optional[GetResult]:
+        # Get all the items in the collection.
+        collection = self.client.get_collection(name=collection_name)
+        if collection:
+            result = collection.get()
+            return GetResult(
+                **{
+                    "ids": [result["ids"]],
+                    "documents": [result["documents"]],
+                    "metadatas": [result["metadatas"]],
+                }
+            )
+        return None
+
+    def insert(self, collection_name: str, items: list[VectorItem]):
+        # Insert the items into the collection, if the collection does not exist, it will be created.
+        collection = self.client.get_or_create_collection(name=collection_name)
+
+        ids = [item["id"] for item in items]
+        documents = [item["text"] for item in items]
+        embeddings = [item["vector"] for item in items]
+        metadatas = [item["metadata"] for item in items]
+
+        for batch in create_batches(
+            api=self.client,
+            documents=documents,
+            embeddings=embeddings,
+            ids=ids,
+            metadatas=metadatas,
+        ):
+            collection.add(*batch)
+
+    def upsert(self, collection_name: str, items: list[VectorItem]):
+        # Update the items in the collection, if the items are not present, insert them. If the collection does not exist, it will be created.
+        collection = self.client.get_or_create_collection(name=collection_name)
+
+        ids = [item["id"] for item in items]
+        documents = [item["text"] for item in items]
+        embeddings = [item["vector"] for item in items]
+        metadatas = [item["metadata"] for item in items]
+
+        collection.upsert(
+            ids=ids, documents=documents, embeddings=embeddings, metadatas=metadatas
+        )
+
+    def delete(self, collection_name: str, ids: list[str]):
+        # Delete the items from the collection based on the ids.
+        collection = self.client.get_collection(name=collection_name)
+        if collection:
+            collection.delete(ids=ids)
+
+    def reset(self):
+        # Resets the database. This will delete all collections and item entries.
+        return self.client.reset()

backend/open_webui/apps/rag/vector/dbs/milvus.py

@@ -0,0 +1,205 @@
+from pymilvus import MilvusClient as Client
+from pymilvus import FieldSchema, DataType
+import json
+
+from typing import Optional
+
+from open_webui.apps.rag.vector.main import VectorItem, SearchResult, GetResult
+from open_webui.config import (
+    MILVUS_URI,
+)
+
+
+class MilvusClient:
+    def __init__(self):
+        self.collection_prefix = "open_webui"
+        self.client = Client(uri=MILVUS_URI)
+
+    def _result_to_get_result(self, result) -> GetResult:
+        print(result)
+
+        ids = []
+        documents = []
+        metadatas = []
+
+        for match in result:
+            _ids = []
+            _documents = []
+            _metadatas = []
+
+            for item in match:
+                _ids.append(item.get("id"))
+                _documents.append(item.get("data", {}).get("text"))
+                _metadatas.append(item.get("metadata"))
+
+            ids.append(_ids)
+            documents.append(_documents)
+            metadatas.append(_metadatas)
+
+        return GetResult(
+            **{
+                "ids": ids,
+                "documents": documents,
+                "metadatas": metadatas,
+            }
+        )
+
+    def _result_to_search_result(self, result) -> SearchResult:
+        print(result)
+
+        ids = []
+        distances = []
+        documents = []
+        metadatas = []
+
+        for match in result:
+            _ids = []
+            _distances = []
+            _documents = []
+            _metadatas = []
+
+            for item in match:
+                _ids.append(item.get("id"))
+                _distances.append(item.get("distance"))
+                _documents.append(item.get("entity", {}).get("data", {}).get("text"))
+                _metadatas.append(item.get("entity", {}).get("metadata"))
+
+            ids.append(_ids)
+            distances.append(_distances)
+            documents.append(_documents)
+            metadatas.append(_metadatas)
+
+        return SearchResult(
+            **{
+                "ids": ids,
+                "distances": distances,
+                "documents": documents,
+                "metadatas": metadatas,
+            }
+        )
+
+    def _create_collection(self, collection_name: str, dimension: int):
+        schema = self.client.create_schema(
+            auto_id=False,
+            enable_dynamic_field=True,
+        )
+        schema.add_field(
+            field_name="id",
+            datatype=DataType.VARCHAR,
+            is_primary=True,
+            max_length=65535,
+        )
+        schema.add_field(
+            field_name="vector",
+            datatype=DataType.FLOAT_VECTOR,
+            dim=dimension,
+            description="vector",
+        )
+        schema.add_field(field_name="data", datatype=DataType.JSON, description="data")
+        schema.add_field(
+            field_name="metadata", datatype=DataType.JSON, description="metadata"
+        )
+
+        index_params = self.client.prepare_index_params()
+        index_params.add_index(
+            field_name="vector", index_type="HNSW", metric_type="COSINE", params={}
+        )
+
+        self.client.create_collection(
+            collection_name=f"{self.collection_prefix}_{collection_name}",
+            schema=schema,
+            index_params=index_params,
+        )
+
+    def has_collection(self, collection_name: str) -> bool:
+        # Check if the collection exists based on the collection name.
+        return self.client.has_collection(
+            collection_name=f"{self.collection_prefix}_{collection_name}"
+        )
+
+    def delete_collection(self, collection_name: str):
+        # Delete the collection based on the collection name.
+        return self.client.drop_collection(
+            collection_name=f"{self.collection_prefix}_{collection_name}"
+        )
+
+    def search(
+        self, collection_name: str, vectors: list[list[float | int]], limit: int
+    ) -> Optional[SearchResult]:
+        # Search for the nearest neighbor items based on the vectors and return 'limit' number of results.
+        result = self.client.search(
+            collection_name=f"{self.collection_prefix}_{collection_name}",
+            data=vectors,
+            limit=limit,
+            output_fields=["data", "metadata"],
+        )
+
+        return self._result_to_search_result(result)
+
+    def get(self, collection_name: str) -> Optional[GetResult]:
+        # Get all the items in the collection.
+        result = self.client.query(
+            collection_name=f"{self.collection_prefix}_{collection_name}",
+            filter='id != ""',
+        )
+        return self._result_to_get_result([result])
+
+    def insert(self, collection_name: str, items: list[VectorItem]):
+        # Insert the items into the collection, if the collection does not exist, it will be created.
+        if not self.client.has_collection(
+            collection_name=f"{self.collection_prefix}_{collection_name}"
+        ):
+            self._create_collection(
+                collection_name=collection_name, dimension=len(items[0]["vector"])
+            )
+
+        return self.client.insert(
+            collection_name=f"{self.collection_prefix}_{collection_name}",
+            data=[
+                {
+                    "id": item["id"],
+                    "vector": item["vector"],
+                    "data": {"text": item["text"]},
+                    "metadata": item["metadata"],
+                }
+                for item in items
+            ],
+        )
+
+    def upsert(self, collection_name: str, items: list[VectorItem]):
+        # Update the items in the collection, if the items are not present, insert them. If the collection does not exist, it will be created.
+        if not self.client.has_collection(
+            collection_name=f"{self.collection_prefix}_{collection_name}"
+        ):
+            self._create_collection(
+                collection_name=collection_name, dimension=len(items[0]["vector"])
+            )
+
+        return self.client.upsert(
+            collection_name=f"{self.collection_prefix}_{collection_name}",
+            data=[
+                {
+                    "id": item["id"],
+                    "vector": item["vector"],
+                    "data": {"text": item["text"]},
+                    "metadata": item["metadata"],
+                }
+                for item in items
+            ],
+        )
+
+    def delete(self, collection_name: str, ids: list[str]):
+        # Delete the items from the collection based on the ids.
+
+        return self.client.delete(
+            collection_name=f"{self.collection_prefix}_{collection_name}",
+            ids=ids,
+        )
+
+    def reset(self):
+        # Resets the database. This will delete all collections and item entries.
+
+        collection_names = self.client.list_collections()
+        for collection_name in collection_names:
+            if collection_name.startswith(self.collection_prefix):
+                self.client.drop_collection(collection_name=collection_name)

backend/open_webui/apps/rag/vector/main.py

@@ -0,0 +1,19 @@
+from pydantic import BaseModel
+from typing import Optional, List, Any
+
+
+class VectorItem(BaseModel):
+    id: str
+    text: str
+    vector: List[float | int]
+    metadata: Any
+
+
+class GetResult(BaseModel):
+    ids: Optional[List[List[str]]]
+    documents: Optional[List[List[str]]]
+    metadatas: Optional[List[List[Any]]]
+
+
+class SearchResult(GetResult):
+    distances: Optional[List[List[float | int]]]

backend/open_webui/apps/socket/main.py

@@ -1,11 +1,39 @@
-import socketio
 import asyncio
 
+import socketio
+from open_webui.apps.webui.models.users import Users
+from open_webui.env import (
+    ENABLE_WEBSOCKET_SUPPORT,
+    WEBSOCKET_MANAGER,
+    WEBSOCKET_REDIS_URL,
+)
+from open_webui.utils.utils import decode_token
+
+
+if WEBSOCKET_MANAGER == "redis":
+    mgr = socketio.AsyncRedisManager(WEBSOCKET_REDIS_URL)
+    sio = socketio.AsyncServer(
+        cors_allowed_origins=[],
+        async_mode="asgi",
+        transports=(
+            ["polling", "websocket"] if ENABLE_WEBSOCKET_SUPPORT else ["polling"]
+        ),
+        allow_upgrades=ENABLE_WEBSOCKET_SUPPORT,
+        always_connect=True,
+        client_manager=mgr,
+    )
+else:
+    sio = socketio.AsyncServer(
+        cors_allowed_origins=[],
+        async_mode="asgi",
+        transports=(
+            ["polling", "websocket"] if ENABLE_WEBSOCKET_SUPPORT else ["polling"]
+        ),
+        allow_upgrades=ENABLE_WEBSOCKET_SUPPORT,
+        always_connect=True,
+    )
 
-from apps.webui.models.users import Users
-from utils.utils import decode_token
 
-sio = socketio.AsyncServer(cors_allowed_origins=[], async_mode="asgi")
 app = socketio.ASGIApp(sio, socketio_path="/ws/socket.io")
 
 # Dictionary to maintain the user pool
@@ -33,7 +61,7 @@ async def connect(sid, environ, auth):
             else:
                 USER_POOL[user.id] = [sid]
 
-            print(f"user {user.name}({user.id}) connected with session ID {sid}")
+            # print(f"user {user.name}({user.id}) connected with session ID {sid}")
 
             await sio.emit("user-count", {"count": len(set(USER_POOL))})
             await sio.emit("usage", {"models": get_models_in_use()})
@@ -41,7 +69,7 @@ async def connect(sid, environ, auth):
 
 @sio.on("user-join")
 async def user_join(sid, data):
-    print("user-join", sid, data)
+    # print("user-join", sid, data)
 
     auth = data["auth"] if "auth" in data else None
     if not auth or "token" not in auth:
@@ -61,7 +89,7 @@ async def user_join(sid, data):
     else:
         USER_POOL[user.id] = [sid]
 
-    print(f"user {user.name}({user.id}) connected with session ID {sid}")
+    # print(f"user {user.name}({user.id}) connected with session ID {sid}")
 
     await sio.emit("user-count", {"count": len(set(USER_POOL))})
 
@@ -110,7 +138,7 @@ async def remove_after_timeout(sid, model_id):
     try:
         await asyncio.sleep(TIMEOUT_DURATION)
         if model_id in USAGE_POOL:
-            print(USAGE_POOL[model_id]["sids"])
+            # print(USAGE_POOL[model_id]["sids"])
             USAGE_POOL[model_id]["sids"].remove(sid)
             USAGE_POOL[model_id]["sids"] = list(set(USAGE_POOL[model_id]["sids"]))
 
@@ -137,7 +165,8 @@ async def disconnect(sid):
 
         await sio.emit("user-count", {"count": len(USER_POOL)})
     else:
-        print(f"Unknown session ID {sid} disconnected")
+        pass
+        # print(f"Unknown session ID {sid} disconnected")
 
 
 def get_event_emitter(request_info):

backend/open_webui/apps/webui/internal/db.py

@@ -1,20 +1,16 @@
-import os
-import logging
 import json
+import logging
 from contextlib import contextmanager
+from typing import Any, Optional
 
+from open_webui.apps.webui.internal.wrappers import register_connection
+from open_webui.env import OPEN_WEBUI_DIR, DATABASE_URL, SRC_LOG_LEVELS
 from peewee_migrate import Router
-from apps.webui.internal.wrappers import register_connection
-
-from typing import Optional, Any
-from typing_extensions import Self
-
-from sqlalchemy import create_engine, types, Dialect
+from sqlalchemy import Dialect, create_engine, types
 from sqlalchemy.ext.declarative import declarative_base
-from sqlalchemy.orm import sessionmaker, scoped_session
+from sqlalchemy.orm import scoped_session, sessionmaker
 from sqlalchemy.sql.type_api import _T
-
-from config import SRC_LOG_LEVELS, DATA_DIR, DATABASE_URL, BACKEND_DIR
+from typing_extensions import Self
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["DB"])
@@ -42,34 +38,21 @@ class JSONField(types.TypeDecorator):
             return json.loads(value)
 
 
-# Check if the file exists
-if os.path.exists(f"{DATA_DIR}/ollama.db"):
-    # Rename the file
-    os.rename(f"{DATA_DIR}/ollama.db", f"{DATA_DIR}/webui.db")
-    log.info("Database migrated from Ollama-WebUI successfully.")
-else:
-    pass
-
-
 # Workaround to handle the peewee migration
 # This is required to ensure the peewee migration is handled before the alembic migration
 def handle_peewee_migration(DATABASE_URL):
+    # db = None
     try:
-        # Replace the postgresql:// with postgres:// and %40 with @ in the DATABASE_URL
-        db = register_connection(
-            DATABASE_URL.replace("postgresql://", "postgres://").replace("%40", "@")
-        )
-        migrate_dir = BACKEND_DIR / "apps" / "webui" / "internal" / "migrations"
+        # Replace the postgresql:// with postgres:// to handle the peewee migration
+        db = register_connection(DATABASE_URL.replace("postgresql://", "postgres://"))
+        migrate_dir = OPEN_WEBUI_DIR / "apps" / "webui" / "internal" / "migrations"
         router = Router(db, logger=log, migrate_dir=migrate_dir)
         router.run()
         db.close()
 
-        # check if db connection has been closed
-
     except Exception as e:
         log.error(f"Failed to initialize the database connection: {e}")
         raise
-
     finally:
         # Properly closing the database connection
         if db and not db.is_closed():
@@ -98,7 +81,6 @@ Base = declarative_base()
 Session = scoped_session(SessionLocal)
 
 
-# Dependency
 def get_session():
     db = SessionLocal()
     try:

backend/open_webui/apps/webui/internal/migrations/010_migrate_modelfiles_to_models.py

@@ -30,7 +30,7 @@ import peewee as pw
 from peewee_migrate import Migrator
 import json
 
-from utils.misc import parse_ollama_modelfile
+from open_webui.utils.misc import parse_ollama_modelfile
 
 with suppress(ImportError):
     import playhouse.postgres_ext as pw_pext

backend/open_webui/apps/webui/internal/wrappers.py

@@ -1,13 +1,13 @@
-from contextvars import ContextVar
-from peewee import *
-from peewee import PostgresqlDatabase, InterfaceError as PeeWeeInterfaceError
-
 import logging
+from contextvars import ContextVar
+
+from open_webui.env import SRC_LOG_LEVELS
+from peewee import *
+from peewee import InterfaceError as PeeWeeInterfaceError
+from peewee import PostgresqlDatabase
 from playhouse.db_url import connect, parse
 from playhouse.shortcuts import ReconnectMixin
 
-from config import SRC_LOG_LEVELS
-
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["DB"])
 
@@ -43,7 +43,7 @@ class ReconnectingPostgresqlDatabase(CustomReconnectMixin, PostgresqlDatabase):
 
 
 def register_connection(db_url):
-    db = connect(db_url)
+    db = connect(db_url, unquote_password=True)
     if isinstance(db, PostgresqlDatabase):
         # Enable autoconnect for SQLite databases, managed by Peewee
         db.autoconnect = True
@@ -51,16 +51,10 @@ def register_connection(db_url):
         log.info("Connected to PostgreSQL database")
 
         # Get the connection details
-        connection = parse(db_url)
+        connection = parse(db_url, unquote_password=True)
 
         # Use our custom database class that supports reconnection
-        db = ReconnectingPostgresqlDatabase(
-            connection["database"],
-            user=connection["user"],
-            password=connection["password"],
-            host=connection["host"],
-            port=connection["port"],
-        )
+        db = ReconnectingPostgresqlDatabase(**connection)
         db.connect(reuse_if_open=True)
     elif isinstance(db, SqliteDatabase):
         # Enable autoconnect for SQLite databases, managed by Peewee

backend/open_webui/apps/webui/main.py

@@ -1,64 +1,70 @@
-from fastapi import FastAPI
-from fastapi.responses import StreamingResponse
-from fastapi.middleware.cors import CORSMiddleware
-from apps.webui.routers import (
+import inspect
+import json
+import logging
+from typing import AsyncGenerator, Generator, Iterator
+
+from open_webui.apps.socket.main import get_event_call, get_event_emitter
+from open_webui.apps.webui.models.functions import Functions
+from open_webui.apps.webui.models.models import Models
+from open_webui.apps.webui.routers import (
     auths,
-    users,
     chats,
-    documents,
-    tools,
-    models,
-    prompts,
     configs,
-    memories,
-    utils,
+    documents,
     files,
     functions,
+    memories,
+    models,
+    prompts,
+    tools,
+    users,
+    utils,
 )
-from apps.webui.models.functions import Functions
-from apps.webui.models.models import Models
-from apps.webui.utils import load_function_module_by_id
-
-from utils.misc import (
+from open_webui.apps.webui.utils import load_function_module_by_id
+from open_webui.config import (
+    ADMIN_EMAIL,
+    CORS_ALLOW_ORIGIN,
+    DEFAULT_MODELS,
+    DEFAULT_PROMPT_SUGGESTIONS,
+    DEFAULT_USER_ROLE,
+    ENABLE_COMMUNITY_SHARING,
+    ENABLE_LOGIN_FORM,
+    ENABLE_MESSAGE_RATING,
+    ENABLE_SIGNUP,
+    JWT_EXPIRES_IN,
+    OAUTH_EMAIL_CLAIM,
+    OAUTH_PICTURE_CLAIM,
+    OAUTH_USERNAME_CLAIM,
+    SHOW_ADMIN_DETAILS,
+    USER_PERMISSIONS,
+    WEBHOOK_URL,
+    WEBUI_AUTH,
+    WEBUI_BANNERS,
+    AppConfig,
+)
+from open_webui.env import (
+    WEBUI_AUTH_TRUSTED_EMAIL_HEADER,
+    WEBUI_AUTH_TRUSTED_NAME_HEADER,
+)
+from fastapi import FastAPI
+from fastapi.middleware.cors import CORSMiddleware
+from fastapi.responses import StreamingResponse
+from pydantic import BaseModel
+from open_webui.utils.misc import (
     openai_chat_chunk_message_template,
     openai_chat_completion_message_template,
-    apply_model_params_to_body,
+)
+from open_webui.utils.payload import (
+    apply_model_params_to_body_openai,
     apply_model_system_prompt_to_body,
 )
 
 
-from config import (
-    SHOW_ADMIN_DETAILS,
-    ADMIN_EMAIL,
-    WEBUI_AUTH,
-    DEFAULT_MODELS,
-    DEFAULT_PROMPT_SUGGESTIONS,
-    DEFAULT_USER_ROLE,
-    ENABLE_SIGNUP,
-    ENABLE_LOGIN_FORM,
-    USER_PERMISSIONS,
-    WEBHOOK_URL,
-    WEBUI_AUTH_TRUSTED_EMAIL_HEADER,
-    WEBUI_AUTH_TRUSTED_NAME_HEADER,
-    JWT_EXPIRES_IN,
-    WEBUI_BANNERS,
-    ENABLE_COMMUNITY_SHARING,
-    AppConfig,
-    OAUTH_USERNAME_CLAIM,
-    OAUTH_PICTURE_CLAIM,
-)
-
-from apps.socket.main import get_event_call, get_event_emitter
-
-import inspect
-import json
-
-from typing import Iterator, Generator, AsyncGenerator
-from pydantic import BaseModel
+from open_webui.utils.tools import get_tools
 
 app = FastAPI()
 
-origins = ["*"]
+log = logging.getLogger(__name__)
 
 app.state.config = AppConfig()
 
@@ -81,9 +87,11 @@ app.state.config.WEBHOOK_URL = WEBHOOK_URL
 app.state.config.BANNERS = WEBUI_BANNERS
 
 app.state.config.ENABLE_COMMUNITY_SHARING = ENABLE_COMMUNITY_SHARING
+app.state.config.ENABLE_MESSAGE_RATING = ENABLE_MESSAGE_RATING
 
 app.state.config.OAUTH_USERNAME_CLAIM = OAUTH_USERNAME_CLAIM
 app.state.config.OAUTH_PICTURE_CLAIM = OAUTH_PICTURE_CLAIM
+app.state.config.OAUTH_EMAIL_CLAIM = OAUTH_EMAIL_CLAIM
 
 app.state.MODELS = {}
 app.state.TOOLS = {}
@@ -91,7 +99,7 @@ app.state.FUNCTIONS = {}
 
 app.add_middleware(
     CORSMiddleware,
-    allow_origins=origins,
+    allow_origins=CORS_ALLOW_ORIGIN,
     allow_credentials=True,
     allow_methods=["*"],
     allow_headers=["*"],
@@ -148,29 +156,33 @@ async def get_pipe_models():
 
         # Check if function is a manifold
         if hasattr(function_module, "pipes"):
-            manifold_pipes = []
+            sub_pipes = []
 
             # Check if pipes is a function or a list
-            if callable(function_module.pipes):
-                manifold_pipes = function_module.pipes()
-            else:
-                manifold_pipes = function_module.pipes
 
-            for p in manifold_pipes:
-                manifold_pipe_id = f'{pipe.id}.{p["id"]}'
-                manifold_pipe_name = p["name"]
+            try:
+                if callable(function_module.pipes):
+                    sub_pipes = function_module.pipes()
+                else:
+                    sub_pipes = function_module.pipes
+            except Exception as e:
+                log.exception(e)
+                sub_pipes = []
+
+            print(sub_pipes)
+
+            for p in sub_pipes:
+                sub_pipe_id = f'{pipe.id}.{p["id"]}'
+                sub_pipe_name = p["name"]
 
                 if hasattr(function_module, "name"):
-                    manifold_pipe_name = f"{function_module.name}{manifold_pipe_name}"
+                    sub_pipe_name = f"{function_module.name}{sub_pipe_name}"
 
                 pipe_flag = {"type": pipe.type}
-                if hasattr(function_module, "ChatValves"):
-                    pipe_flag["valves_spec"] = function_module.ChatValves.schema()
-
                 pipe_models.append(
                     {
-                        "id": manifold_pipe_id,
-                        "name": manifold_pipe_name,
+                        "id": sub_pipe_id,
+                        "name": sub_pipe_name,
                         "object": "model",
                         "created": pipe.created_at,
                         "owned_by": "openai",
@@ -179,8 +191,6 @@ async def get_pipe_models():
                 )
         else:
             pipe_flag = {"type": "pipe"}
-            if hasattr(function_module, "ChatValves"):
-                pipe_flag["valves_spec"] = function_module.ChatValves.schema()
 
             pipe_models.append(
                 {
@@ -239,40 +249,40 @@ def get_pipe_id(form_data: dict) -> str:
     return pipe_id
 
 
-def get_function_params(function_module, form_data, user, extra_params={}):
+def get_function_params(function_module, form_data, user, extra_params=None):
+    if extra_params is None:
+        extra_params = {}
+
     pipe_id = get_pipe_id(form_data)
+
     # Get the signature of the function
     sig = inspect.signature(function_module.pipe)
-    params = {"body": form_data}
-
-    for key, value in extra_params.items():
-        if key in sig.parameters:
-            params[key] = value
-
-    if "__user__" in sig.parameters:
-        __user__ = {
-            "id": user.id,
-            "email": user.email,
-            "name": user.name,
-            "role": user.role,
-        }
+    params = {"body": form_data} | {
+        k: v for k, v in extra_params.items() if k in sig.parameters
+    }
 
+    if "__user__" in params and hasattr(function_module, "UserValves"):
+        user_valves = Functions.get_user_valves_by_id_and_user_id(pipe_id, user.id)
         try:
-            if hasattr(function_module, "UserValves"):
-                __user__["valves"] = function_module.UserValves(
-                    **Functions.get_user_valves_by_id_and_user_id(pipe_id, user.id)
-                )
+            params["__user__"]["valves"] = function_module.UserValves(**user_valves)
         except Exception as e:
-            print(e)
+            log.exception(e)
+            params["__user__"]["valves"] = function_module.UserValves()
 
-        params["__user__"] = __user__
     return params
 
 
 async def generate_function_chat_completion(form_data, user):
     model_id = form_data.get("model")
     model_info = Models.get_model_by_id(model_id)
-    metadata = form_data.pop("metadata", None)
+
+    metadata = form_data.pop("metadata", {})
+
+    files = metadata.get("files", [])
+    tool_ids = metadata.get("tool_ids", [])
+    # Check if tool_ids is None
+    if tool_ids is None:
+        tool_ids = []
 
     __event_emitter__ = None
     __event_call__ = None
@@ -284,28 +294,43 @@ async def generate_function_chat_completion(form_data, user):
             __event_call__ = get_event_call(metadata)
         __task__ = metadata.get("task", None)
 
+    extra_params = {
+        "__event_emitter__": __event_emitter__,
+        "__event_call__": __event_call__,
+        "__task__": __task__,
+        "__files__": files,
+        "__user__": {
+            "id": user.id,
+            "email": user.email,
+            "name": user.name,
+            "role": user.role,
+        },
+    }
+    extra_params["__tools__"] = get_tools(
+        app,
+        tool_ids,
+        user,
+        {
+            **extra_params,
+            "__model__": app.state.MODELS[form_data["model"]],
+            "__messages__": form_data["messages"],
+            "__files__": files,
+        },
+    )
+
     if model_info:
         if model_info.base_model_id:
             form_data["model"] = model_info.base_model_id
 
         params = model_info.params.model_dump()
-        form_data = apply_model_params_to_body(params, form_data)
+        form_data = apply_model_params_to_body_openai(params, form_data)
         form_data = apply_model_system_prompt_to_body(params, form_data, user)
 
     pipe_id = get_pipe_id(form_data)
     function_module = get_function_module(pipe_id)
 
     pipe = function_module.pipe
-    params = get_function_params(
-        function_module,
-        form_data,
-        user,
-        {
-            "__event_emitter__": __event_emitter__,
-            "__event_call__": __event_call__,
-            "__task__": __task__,
-        },
-    )
+    params = get_function_params(function_module, form_data, user, extra_params)
 
     if form_data["stream"]:
 

backend/open_webui/apps/webui/models/auths.py

@@ -1,15 +1,13 @@
-from pydantic import BaseModel
-from typing import Optional
-import uuid
 import logging
-from sqlalchemy import String, Column, Boolean, Text
+import uuid
+from typing import Optional
 
-from apps.webui.models.users import UserModel, Users
-from utils.utils import verify_password
-
-from apps.webui.internal.db import Base, get_db
-
-from config import SRC_LOG_LEVELS
+from open_webui.apps.webui.internal.db import Base, get_db
+from open_webui.apps.webui.models.users import UserModel, Users
+from open_webui.env import SRC_LOG_LEVELS
+from pydantic import BaseModel
+from sqlalchemy import Boolean, Column, String, Text
+from open_webui.utils.utils import verify_password
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["MODELS"])
@@ -92,7 +90,6 @@ class AddUserForm(SignupForm):
 
 
 class AuthsTable:
-
     def insert_new_auth(
         self,
         email: str,
@@ -103,7 +100,6 @@ class AuthsTable:
         oauth_sub: Optional[str] = None,
     ) -> Optional[UserModel]:
         with get_db() as db:
-
             log.info("insert_new_auth")
 
             id = str(uuid.uuid4())
@@ -130,7 +126,6 @@ class AuthsTable:
         log.info(f"authenticate_user: {email}")
         try:
             with get_db() as db:
-
                 auth = db.query(Auth).filter_by(email=email, active=True).first()
                 if auth:
                     if verify_password(password, auth.password):
@@ -140,7 +135,7 @@ class AuthsTable:
                         return None
                 else:
                     return None
-        except:
+        except Exception:
             return None
 
     def authenticate_user_by_api_key(self, api_key: str) -> Optional[UserModel]:
@@ -152,7 +147,7 @@ class AuthsTable:
         try:
             user = Users.get_user_by_api_key(api_key)
             return user if user else None
-        except:
+        except Exception:
             return False
 
     def authenticate_user_by_trusted_header(self, email: str) -> Optional[UserModel]:
@@ -163,7 +158,7 @@ class AuthsTable:
                 if auth:
                     user = Users.get_user_by_id(auth.id)
                     return user
-        except:
+        except Exception:
             return None
 
     def update_user_password_by_id(self, id: str, new_password: str) -> bool:
@@ -174,7 +169,7 @@ class AuthsTable:
                 )
                 db.commit()
                 return True if result == 1 else False
-        except:
+        except Exception:
             return False
 
     def update_email_by_id(self, id: str, email: str) -> bool:
@@ -183,13 +178,12 @@ class AuthsTable:
                 result = db.query(Auth).filter_by(id=id).update({"email": email})
                 db.commit()
                 return True if result == 1 else False
-        except:
+        except Exception:
             return False
 
     def delete_auth_by_id(self, id: str) -> bool:
         try:
             with get_db() as db:
-
                 # Delete User
                 result = Users.delete_user_by_id(id)
 
@@ -200,7 +194,7 @@ class AuthsTable:
                     return True
                 else:
                     return False
-        except:
+        except Exception:
             return False
 
 

backend/open_webui/apps/webui/models/chats.py

@@ -1,14 +1,11 @@
-from pydantic import BaseModel, ConfigDict
-from typing import List, Union, Optional
-
 import json
-import uuid
 import time
+import uuid
+from typing import Optional
 
-from sqlalchemy import Column, String, BigInteger, Boolean, Text
-
-from apps.webui.internal.db import Base, get_db
-
+from open_webui.apps.webui.internal.db import Base, get_db
+from pydantic import BaseModel, ConfigDict
+from sqlalchemy import BigInteger, Boolean, Column, String, Text
 
 ####################
 # Chat DB Schema
@@ -77,10 +74,8 @@ class ChatTitleIdResponse(BaseModel):
 
 
 class ChatTable:
-
     def insert_new_chat(self, user_id: str, form_data: ChatForm) -> Optional[ChatModel]:
         with get_db() as db:
-
             id = str(uuid.uuid4())
             chat = ChatModel(
                 **{
@@ -106,7 +101,6 @@ class ChatTable:
     def update_chat_by_id(self, id: str, chat: dict) -> Optional[ChatModel]:
         try:
             with get_db() as db:
-
                 chat_obj = db.get(Chat, id)
                 chat_obj.chat = json.dumps(chat)
                 chat_obj.title = chat["title"] if "title" in chat else "New Chat"
@@ -115,12 +109,11 @@ class ChatTable:
                 db.refresh(chat_obj)
 
                 return ChatModel.model_validate(chat_obj)
-        except Exception as e:
+        except Exception:
             return None
 
     def insert_shared_chat_by_chat_id(self, chat_id: str) -> Optional[ChatModel]:
         with get_db() as db:
-
             # Get the existing chat to share
             chat = db.get(Chat, chat_id)
             # Check if the chat is already shared
@@ -154,7 +147,6 @@ class ChatTable:
     def update_shared_chat_by_chat_id(self, chat_id: str) -> Optional[ChatModel]:
         try:
             with get_db() as db:
-
                 print("update_shared_chat_by_id")
                 chat = db.get(Chat, chat_id)
                 print(chat)
@@ -164,18 +156,17 @@ class ChatTable:
                 db.refresh(chat)
 
                 return self.get_chat_by_id(chat.share_id)
-        except:
+        except Exception:
             return None
 
     def delete_shared_chat_by_chat_id(self, chat_id: str) -> bool:
         try:
             with get_db() as db:
-
                 db.query(Chat).filter_by(user_id=f"shared-{chat_id}").delete()
                 db.commit()
 
                 return True
-        except:
+        except Exception:
             return False
 
     def update_chat_share_id_by_id(
@@ -183,25 +174,23 @@ class ChatTable:
     ) -> Optional[ChatModel]:
         try:
             with get_db() as db:
-
                 chat = db.get(Chat, id)
                 chat.share_id = share_id
                 db.commit()
                 db.refresh(chat)
                 return ChatModel.model_validate(chat)
-        except:
+        except Exception:
             return None
 
     def toggle_chat_archive_by_id(self, id: str) -> Optional[ChatModel]:
         try:
             with get_db() as db:
-
                 chat = db.get(Chat, id)
                 chat.archived = not chat.archived
                 db.commit()
                 db.refresh(chat)
                 return ChatModel.model_validate(chat)
-        except:
+        except Exception:
             return None
 
     def archive_all_chats_by_user_id(self, user_id: str) -> bool:
@@ -210,14 +199,13 @@ class ChatTable:
                 db.query(Chat).filter_by(user_id=user_id).update({"archived": True})
                 db.commit()
                 return True
-        except:
+        except Exception:
             return False
 
     def get_archived_chat_list_by_user_id(
         self, user_id: str, skip: int = 0, limit: int = 50
-    ) -> List[ChatModel]:
+    ) -> list[ChatModel]:
         with get_db() as db:
-
             all_chats = (
                 db.query(Chat)
                 .filter_by(user_id=user_id, archived=True)
@@ -233,7 +221,7 @@ class ChatTable:
         include_archived: bool = False,
         skip: int = 0,
         limit: int = 50,
-    ) -> List[ChatModel]:
+    ) -> list[ChatModel]:
         with get_db() as db:
             query = db.query(Chat).filter_by(user_id=user_id)
             if not include_archived:
@@ -249,22 +237,25 @@ class ChatTable:
         self,
         user_id: str,
         include_archived: bool = False,
-        skip: int = 0,
-        limit: int = -1,
-    ) -> List[ChatTitleIdResponse]:
+        skip: Optional[int] = None,
+        limit: Optional[int] = None,
+    ) -> list[ChatTitleIdResponse]:
         with get_db() as db:
             query = db.query(Chat).filter_by(user_id=user_id)
             if not include_archived:
                 query = query.filter_by(archived=False)
 
-            all_chats = (
-                query.order_by(Chat.updated_at.desc())
-                # limit cols
-                .with_entities(Chat.id, Chat.title, Chat.updated_at, Chat.created_at)
-                .limit(limit)
-                .offset(skip)
-                .all()
+            query = query.order_by(Chat.updated_at.desc()).with_entities(
+                Chat.id, Chat.title, Chat.updated_at, Chat.created_at
             )
+
+            if limit:
+                query = query.limit(limit)
+            if skip:
+                query = query.offset(skip)
+
+            all_chats = query.all()
+
             # result has to be destrctured from sqlalchemy `row` and mapped to a dict since the `ChatModel`is not the returned dataclass.
             return [
                 ChatTitleIdResponse.model_validate(
@@ -279,8 +270,8 @@ class ChatTable:
             ]
 
     def get_chat_list_by_chat_ids(
-        self, chat_ids: List[str], skip: int = 0, limit: int = 50
-    ) -> List[ChatModel]:
+        self, chat_ids: list[str], skip: int = 0, limit: int = 50
+    ) -> list[ChatModel]:
         with get_db() as db:
             all_chats = (
                 db.query(Chat)
@@ -294,37 +285,33 @@ class ChatTable:
     def get_chat_by_id(self, id: str) -> Optional[ChatModel]:
         try:
             with get_db() as db:
-
                 chat = db.get(Chat, id)
                 return ChatModel.model_validate(chat)
-        except:
+        except Exception:
             return None
 
     def get_chat_by_share_id(self, id: str) -> Optional[ChatModel]:
         try:
             with get_db() as db:
-
                 chat = db.query(Chat).filter_by(share_id=id).first()
 
                 if chat:
                     return self.get_chat_by_id(id)
                 else:
                     return None
-        except Exception as e:
+        except Exception:
             return None
 
     def get_chat_by_id_and_user_id(self, id: str, user_id: str) -> Optional[ChatModel]:
         try:
             with get_db() as db:
-
                 chat = db.query(Chat).filter_by(id=id, user_id=user_id).first()
                 return ChatModel.model_validate(chat)
-        except:
+        except Exception:
             return None
 
-    def get_chats(self, skip: int = 0, limit: int = 50) -> List[ChatModel]:
+    def get_chats(self, skip: int = 0, limit: int = 50) -> list[ChatModel]:
         with get_db() as db:
-
             all_chats = (
                 db.query(Chat)
                 # .limit(limit).offset(skip)
@@ -332,9 +319,8 @@ class ChatTable:
             )
             return [ChatModel.model_validate(chat) for chat in all_chats]
 
-    def get_chats_by_user_id(self, user_id: str) -> List[ChatModel]:
+    def get_chats_by_user_id(self, user_id: str) -> list[ChatModel]:
         with get_db() as db:
-
             all_chats = (
                 db.query(Chat)
                 .filter_by(user_id=user_id)
@@ -342,9 +328,8 @@ class ChatTable:
             )
             return [ChatModel.model_validate(chat) for chat in all_chats]
 
-    def get_archived_chats_by_user_id(self, user_id: str) -> List[ChatModel]:
+    def get_archived_chats_by_user_id(self, user_id: str) -> list[ChatModel]:
         with get_db() as db:
-
             all_chats = (
                 db.query(Chat)
                 .filter_by(user_id=user_id, archived=True)
@@ -355,44 +340,38 @@ class ChatTable:
     def delete_chat_by_id(self, id: str) -> bool:
         try:
             with get_db() as db:
-
                 db.query(Chat).filter_by(id=id).delete()
                 db.commit()
 
                 return True and self.delete_shared_chat_by_chat_id(id)
-        except:
+        except Exception:
             return False
 
     def delete_chat_by_id_and_user_id(self, id: str, user_id: str) -> bool:
         try:
             with get_db() as db:
-
                 db.query(Chat).filter_by(id=id, user_id=user_id).delete()
                 db.commit()
 
                 return True and self.delete_shared_chat_by_chat_id(id)
-        except:
+        except Exception:
             return False
 
     def delete_chats_by_user_id(self, user_id: str) -> bool:
         try:
-
             with get_db() as db:
-
                 self.delete_shared_chats_by_user_id(user_id)
 
                 db.query(Chat).filter_by(user_id=user_id).delete()
                 db.commit()
 
                 return True
-        except:
+        except Exception:
             return False
 
     def delete_shared_chats_by_user_id(self, user_id: str) -> bool:
         try:
-
             with get_db() as db:
-
                 chats_by_user = db.query(Chat).filter_by(user_id=user_id).all()
                 shared_chat_ids = [f"shared-{chat.id}" for chat in chats_by_user]
 
@@ -400,7 +379,7 @@ class ChatTable:
                 db.commit()
 
                 return True
-        except:
+        except Exception:
             return False
 
 

backend/open_webui/apps/webui/models/documents.py

@@ -1,15 +1,12 @@
-from pydantic import BaseModel, ConfigDict
-from typing import List, Optional
-import time
-import logging
-
-from sqlalchemy import String, Column, BigInteger, Text
-
-from apps.webui.internal.db import Base, get_db
-
 import json
+import logging
+import time
+from typing import Optional
 
-from config import SRC_LOG_LEVELS
+from open_webui.apps.webui.internal.db import Base, get_db
+from open_webui.env import SRC_LOG_LEVELS
+from pydantic import BaseModel, ConfigDict
+from sqlalchemy import BigInteger, Column, String, Text
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["MODELS"])
@@ -70,12 +67,10 @@ class DocumentForm(DocumentUpdateForm):
 
 
 class DocumentsTable:
-
     def insert_new_doc(
         self, user_id: str, form_data: DocumentForm
     ) -> Optional[DocumentModel]:
         with get_db() as db:
-
             document = DocumentModel(
                 **{
                     **form_data.model_dump(),
@@ -93,21 +88,19 @@ class DocumentsTable:
                     return DocumentModel.model_validate(result)
                 else:
                     return None
-            except:
+            except Exception:
                 return None
 
     def get_doc_by_name(self, name: str) -> Optional[DocumentModel]:
         try:
             with get_db() as db:
-
                 document = db.query(Document).filter_by(name=name).first()
                 return DocumentModel.model_validate(document) if document else None
-        except:
+        except Exception:
             return None
 
-    def get_docs(self) -> List[DocumentModel]:
+    def get_docs(self) -> list[DocumentModel]:
         with get_db() as db:
-
             return [
                 DocumentModel.model_validate(doc) for doc in db.query(Document).all()
             ]
@@ -117,7 +110,6 @@ class DocumentsTable:
     ) -> Optional[DocumentModel]:
         try:
             with get_db() as db:
-
                 db.query(Document).filter_by(name=name).update(
                     {
                         "title": form_data.title,
@@ -140,7 +132,6 @@ class DocumentsTable:
             doc_content = {**doc_content, **updated}
 
             with get_db() as db:
-
                 db.query(Document).filter_by(name=name).update(
                     {
                         "content": json.dumps(doc_content),
@@ -156,11 +147,10 @@ class DocumentsTable:
     def delete_doc_by_name(self, name: str) -> bool:
         try:
             with get_db() as db:
-
                 db.query(Document).filter_by(name=name).delete()
                 db.commit()
                 return True
-        except:
+        except Exception:
             return False
 
 

backend/open_webui/apps/webui/models/files.py

@@ -1,15 +1,11 @@
-from pydantic import BaseModel, ConfigDict
-from typing import List, Union, Optional
-import time
 import logging
+import time
+from typing import Optional
 
-from sqlalchemy import Column, String, BigInteger, Text
-
-from apps.webui.internal.db import JSONField, Base, get_db
-
-import json
-
-from config import SRC_LOG_LEVELS
+from open_webui.apps.webui.internal.db import Base, JSONField, get_db
+from open_webui.env import SRC_LOG_LEVELS
+from pydantic import BaseModel, ConfigDict
+from sqlalchemy import BigInteger, Column, String, Text
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["MODELS"])
@@ -59,10 +55,8 @@ class FileForm(BaseModel):
 
 
 class FilesTable:
-
     def insert_new_file(self, user_id: str, form_data: FileForm) -> Optional[FileModel]:
         with get_db() as db:
-
             file = FileModel(
                 **{
                     **form_data.model_dump(),
@@ -86,40 +80,41 @@ class FilesTable:
 
     def get_file_by_id(self, id: str) -> Optional[FileModel]:
         with get_db() as db:
-
             try:
                 file = db.get(File, id)
                 return FileModel.model_validate(file)
-            except:
+            except Exception:
                 return None
 
-    def get_files(self) -> List[FileModel]:
+    def get_files(self) -> list[FileModel]:
         with get_db() as db:
-
             return [FileModel.model_validate(file) for file in db.query(File).all()]
 
-    def delete_file_by_id(self, id: str) -> bool:
-
+    def get_files_by_user_id(self, user_id: str) -> list[FileModel]:
         with get_db() as db:
+            return [
+                FileModel.model_validate(file)
+                for file in db.query(File).filter_by(user_id=user_id).all()
+            ]
 
+    def delete_file_by_id(self, id: str) -> bool:
+        with get_db() as db:
             try:
                 db.query(File).filter_by(id=id).delete()
                 db.commit()
 
                 return True
-            except:
+            except Exception:
                 return False
 
     def delete_all_files(self) -> bool:
-
         with get_db() as db:
-
             try:
                 db.query(File).delete()
                 db.commit()
 
                 return True
-            except:
+            except Exception:
                 return False
 
 

backend/open_webui/apps/webui/models/functions.py

@@ -1,18 +1,12 @@
-from pydantic import BaseModel, ConfigDict
-from typing import List, Union, Optional
-import time
 import logging
+import time
+from typing import Optional
 
-from sqlalchemy import Column, String, Text, BigInteger, Boolean
-
-from apps.webui.internal.db import JSONField, Base, get_db
-from apps.webui.models.users import Users
-
-import json
-import copy
-
-
-from config import SRC_LOG_LEVELS
+from open_webui.apps.webui.internal.db import Base, JSONField, get_db
+from open_webui.apps.webui.models.users import Users
+from open_webui.env import SRC_LOG_LEVELS
+from pydantic import BaseModel, ConfigDict
+from sqlalchemy import BigInteger, Boolean, Column, String, Text
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["MODELS"])
@@ -87,11 +81,9 @@ class FunctionValves(BaseModel):
 
 
 class FunctionsTable:
-
     def insert_new_function(
         self, user_id: str, type: str, form_data: FunctionForm
     ) -> Optional[FunctionModel]:
-
         function = FunctionModel(
             **{
                 **form_data.model_dump(),
@@ -119,15 +111,13 @@ class FunctionsTable:
     def get_function_by_id(self, id: str) -> Optional[FunctionModel]:
         try:
             with get_db() as db:
-
                 function = db.get(Function, id)
                 return FunctionModel.model_validate(function)
-        except:
+        except Exception:
             return None
 
-    def get_functions(self, active_only=False) -> List[FunctionModel]:
+    def get_functions(self, active_only=False) -> list[FunctionModel]:
         with get_db() as db:
-
             if active_only:
                 return [
                     FunctionModel.model_validate(function)
@@ -141,9 +131,8 @@ class FunctionsTable:
 
     def get_functions_by_type(
         self, type: str, active_only=False
-    ) -> List[FunctionModel]:
+    ) -> list[FunctionModel]:
         with get_db() as db:
-
             if active_only:
                 return [
                     FunctionModel.model_validate(function)
@@ -157,9 +146,8 @@ class FunctionsTable:
                     for function in db.query(Function).filter_by(type=type).all()
                 ]
 
-    def get_global_filter_functions(self) -> List[FunctionModel]:
+    def get_global_filter_functions(self) -> list[FunctionModel]:
         with get_db() as db:
-
             return [
                 FunctionModel.model_validate(function)
                 for function in db.query(Function)
@@ -167,7 +155,7 @@ class FunctionsTable:
                 .all()
             ]
 
-    def get_global_action_functions(self) -> List[FunctionModel]:
+    def get_global_action_functions(self) -> list[FunctionModel]:
         with get_db() as db:
             return [
                 FunctionModel.model_validate(function)
@@ -178,7 +166,6 @@ class FunctionsTable:
 
     def get_function_valves_by_id(self, id: str) -> Optional[dict]:
         with get_db() as db:
-
             try:
                 function = db.get(Function, id)
                 return function.valves if function.valves else {}
@@ -190,7 +177,6 @@ class FunctionsTable:
         self, id: str, valves: dict
     ) -> Optional[FunctionValves]:
         with get_db() as db:
-
             try:
                 function = db.get(Function, id)
                 function.valves = valves
@@ -198,13 +184,12 @@ class FunctionsTable:
                 db.commit()
                 db.refresh(function)
                 return self.get_function_by_id(id)
-            except:
+            except Exception:
                 return None
 
     def get_user_valves_by_id_and_user_id(
         self, id: str, user_id: str
     ) -> Optional[dict]:
-
         try:
             user = Users.get_user_by_id(user_id)
             user_settings = user.settings.model_dump() if user.settings else {}
@@ -223,7 +208,6 @@ class FunctionsTable:
     def update_user_valves_by_id_and_user_id(
         self, id: str, user_id: str, valves: dict
     ) -> Optional[dict]:
-
         try:
             user = Users.get_user_by_id(user_id)
             user_settings = user.settings.model_dump() if user.settings else {}
@@ -246,7 +230,6 @@ class FunctionsTable:
 
     def update_function_by_id(self, id: str, updated: dict) -> Optional[FunctionModel]:
         with get_db() as db:
-
             try:
                 db.query(Function).filter_by(id=id).update(
                     {
@@ -256,12 +239,11 @@ class FunctionsTable:
                 )
                 db.commit()
                 return self.get_function_by_id(id)
-            except:
+            except Exception:
                 return None
 
     def deactivate_all_functions(self) -> Optional[bool]:
         with get_db() as db:
-
             try:
                 db.query(Function).update(
                     {
@@ -271,7 +253,7 @@ class FunctionsTable:
                 )
                 db.commit()
                 return True
-            except:
+            except Exception:
                 return None
 
     def delete_function_by_id(self, id: str) -> bool:
@@ -281,7 +263,7 @@ class FunctionsTable:
                 db.commit()
 
                 return True
-            except:
+            except Exception:
                 return False
 
 

backend/open_webui/apps/webui/models/memories.py

@@ -1,12 +1,10 @@
-from pydantic import BaseModel, ConfigDict
-from typing import List, Union, Optional
-
-from sqlalchemy import Column, String, BigInteger, Text
-
-from apps.webui.internal.db import Base, get_db
-
 import time
 import uuid
+from typing import Optional
+
+from open_webui.apps.webui.internal.db import Base, get_db
+from pydantic import BaseModel, ConfigDict
+from sqlalchemy import BigInteger, Column, String, Text
 
 ####################
 # Memory DB Schema
@@ -39,13 +37,11 @@ class MemoryModel(BaseModel):
 
 
 class MemoriesTable:
-
     def insert_new_memory(
         self,
         user_id: str,
         content: str,
     ) -> Optional[MemoryModel]:
-
         with get_db() as db:
             id = str(uuid.uuid4())
 
@@ -73,75 +69,68 @@ class MemoriesTable:
         content: str,
     ) -> Optional[MemoryModel]:
         with get_db() as db:
-
             try:
                 db.query(Memory).filter_by(id=id).update(
                     {"content": content, "updated_at": int(time.time())}
                 )
                 db.commit()
                 return self.get_memory_by_id(id)
-            except:
+            except Exception:
                 return None
 
-    def get_memories(self) -> List[MemoryModel]:
+    def get_memories(self) -> list[MemoryModel]:
         with get_db() as db:
-
             try:
                 memories = db.query(Memory).all()
                 return [MemoryModel.model_validate(memory) for memory in memories]
-            except:
+            except Exception:
                 return None
 
-    def get_memories_by_user_id(self, user_id: str) -> List[MemoryModel]:
+    def get_memories_by_user_id(self, user_id: str) -> list[MemoryModel]:
         with get_db() as db:
-
             try:
                 memories = db.query(Memory).filter_by(user_id=user_id).all()
                 return [MemoryModel.model_validate(memory) for memory in memories]
-            except:
+            except Exception:
                 return None
 
     def get_memory_by_id(self, id: str) -> Optional[MemoryModel]:
         with get_db() as db:
-
             try:
                 memory = db.get(Memory, id)
                 return MemoryModel.model_validate(memory)
-            except:
+            except Exception:
                 return None
 
     def delete_memory_by_id(self, id: str) -> bool:
         with get_db() as db:
-
             try:
                 db.query(Memory).filter_by(id=id).delete()
                 db.commit()
 
                 return True
 
-            except:
+            except Exception:
                 return False
 
     def delete_memories_by_user_id(self, user_id: str) -> bool:
         with get_db() as db:
-
             try:
                 db.query(Memory).filter_by(user_id=user_id).delete()
                 db.commit()
 
                 return True
-            except:
+            except Exception:
                 return False
 
     def delete_memory_by_id_and_user_id(self, id: str, user_id: str) -> bool:
         with get_db() as db:
-
             try:
                 db.query(Memory).filter_by(id=id, user_id=user_id).delete()
                 db.commit()
 
                 return True
-            except:
+            except Exception:
                 return False
 
 

backend/open_webui/apps/webui/models/models.py

@@ -1,14 +1,11 @@
 import logging
-from typing import Optional, List
-
-from pydantic import BaseModel, ConfigDict
-from sqlalchemy import Column, BigInteger, Text
-
-from apps.webui.internal.db import Base, JSONField, get_db
-
-from config import SRC_LOG_LEVELS
-
 import time
+from typing import Optional
+
+from open_webui.apps.webui.internal.db import Base, JSONField, get_db
+from open_webui.env import SRC_LOG_LEVELS
+from pydantic import BaseModel, ConfigDict
+from sqlalchemy import BigInteger, Column, Text
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["MODELS"])
@@ -137,7 +134,7 @@ class ModelsTable:
             print(e)
             return None
 
-    def get_all_models(self) -> List[ModelModel]:
+    def get_all_models(self) -> list[ModelModel]:
         with get_db() as db:
             return [ModelModel.model_validate(model) for model in db.query(Model).all()]
 
@@ -146,7 +143,7 @@ class ModelsTable:
             with get_db() as db:
                 model = db.get(Model, id)
                 return ModelModel.model_validate(model)
-        except:
+        except Exception:
             return None
 
     def update_model_by_id(self, id: str, model: ModelForm) -> Optional[ModelModel]:
@@ -175,7 +172,7 @@ class ModelsTable:
                 db.commit()
 
                 return True
-        except:
+        except Exception:
             return False
 
 

backend/open_webui/apps/webui/models/prompts.py

@@ -1,12 +1,9 @@
-from pydantic import BaseModel, ConfigDict
-from typing import List, Optional
 import time
+from typing import Optional
 
-from sqlalchemy import String, Column, BigInteger, Text
-
-from apps.webui.internal.db import Base, get_db
-
-import json
+from open_webui.apps.webui.internal.db import Base, get_db
+from pydantic import BaseModel, ConfigDict
+from sqlalchemy import BigInteger, Column, String, Text
 
 ####################
 # Prompts DB Schema
@@ -45,7 +42,6 @@ class PromptForm(BaseModel):
 
 
 class PromptsTable:
-
     def insert_new_prompt(
         self, user_id: str, form_data: PromptForm
     ) -> Optional[PromptModel]:
@@ -61,7 +57,6 @@ class PromptsTable:
 
         try:
             with get_db() as db:
-
                 result = Prompt(**prompt.dict())
                 db.add(result)
                 db.commit()
@@ -70,21 +65,19 @@ class PromptsTable:
                     return PromptModel.model_validate(result)
                 else:
                     return None
-        except Exception as e:
+        except Exception:
             return None
 
     def get_prompt_by_command(self, command: str) -> Optional[PromptModel]:
         try:
             with get_db() as db:
-
                 prompt = db.query(Prompt).filter_by(command=command).first()
                 return PromptModel.model_validate(prompt)
-        except:
+        except Exception:
             return None
 
-    def get_prompts(self) -> List[PromptModel]:
+    def get_prompts(self) -> list[PromptModel]:
         with get_db() as db:
-
             return [
                 PromptModel.model_validate(prompt) for prompt in db.query(Prompt).all()
             ]
@@ -94,25 +87,23 @@ class PromptsTable:
     ) -> Optional[PromptModel]:
         try:
             with get_db() as db:
-
                 prompt = db.query(Prompt).filter_by(command=command).first()
                 prompt.title = form_data.title
                 prompt.content = form_data.content
                 prompt.timestamp = int(time.time())
                 db.commit()
                 return PromptModel.model_validate(prompt)
-        except:
+        except Exception:
             return None
 
     def delete_prompt_by_command(self, command: str) -> bool:
         try:
             with get_db() as db:
-
                 db.query(Prompt).filter_by(command=command).delete()
                 db.commit()
 
                 return True
-        except:
+        except Exception:
             return False
 
 

backend/open_webui/apps/webui/models/tags.py

@@ -1,16 +1,12 @@
-from pydantic import BaseModel, ConfigDict
-from typing import List, Optional
-
-import json
-import uuid
-import time
 import logging
+import time
+import uuid
+from typing import Optional
 
-from sqlalchemy import String, Column, BigInteger, Text
-
-from apps.webui.internal.db import Base, get_db
-
-from config import SRC_LOG_LEVELS
+from open_webui.apps.webui.internal.db import Base, get_db
+from open_webui.env import SRC_LOG_LEVELS
+from pydantic import BaseModel, ConfigDict
+from sqlalchemy import BigInteger, Column, String, Text
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["MODELS"])
@@ -69,18 +65,16 @@ class ChatIdTagForm(BaseModel):
 
 
 class TagChatIdsResponse(BaseModel):
-    chat_ids: List[str]
+    chat_ids: list[str]
 
 
 class ChatTagsResponse(BaseModel):
-    tags: List[str]
+    tags: list[str]
 
 
 class TagTable:
-
     def insert_new_tag(self, name: str, user_id: str) -> Optional[TagModel]:
         with get_db() as db:
-
             id = str(uuid.uuid4())
             tag = TagModel(**{"id": id, "user_id": user_id, "name": name})
             try:
@@ -92,7 +86,7 @@ class TagTable:
                     return TagModel.model_validate(result)
                 else:
                     return None
-            except Exception as e:
+            except Exception:
                 return None
 
     def get_tag_by_name_and_user_id(
@@ -102,14 +96,14 @@ class TagTable:
             with get_db() as db:
                 tag = db.query(Tag).filter_by(name=name, user_id=user_id).first()
                 return TagModel.model_validate(tag)
-        except Exception as e:
+        except Exception:
             return None
 
     def add_tag_to_chat(
         self, user_id: str, form_data: ChatIdTagForm
     ) -> Optional[ChatIdTagModel]:
         tag = self.get_tag_by_name_and_user_id(form_data.tag_name, user_id)
-        if tag == None:
+        if tag is None:
             tag = self.insert_new_tag(form_data.tag_name, user_id)
 
         id = str(uuid.uuid4())
@@ -132,10 +126,10 @@ class TagTable:
                     return ChatIdTagModel.model_validate(result)
                 else:
                     return None
-        except:
+        except Exception:
             return None
 
-    def get_tags_by_user_id(self, user_id: str) -> List[TagModel]:
+    def get_tags_by_user_id(self, user_id: str) -> list[TagModel]:
         with get_db() as db:
             tag_names = [
                 chat_id_tag.tag_name
@@ -159,9 +153,8 @@ class TagTable:
 
     def get_tags_by_chat_id_and_user_id(
         self, chat_id: str, user_id: str
-    ) -> List[TagModel]:
+    ) -> list[TagModel]:
         with get_db() as db:
-
             tag_names = [
                 chat_id_tag.tag_name
                 for chat_id_tag in (
@@ -184,9 +177,8 @@ class TagTable:
 
     def get_chat_ids_by_tag_name_and_user_id(
         self, tag_name: str, user_id: str
-    ) -> List[ChatIdTagModel]:
+    ) -> list[ChatIdTagModel]:
         with get_db() as db:
-
             return [
                 ChatIdTagModel.model_validate(chat_id_tag)
                 for chat_id_tag in (
@@ -201,7 +193,6 @@ class TagTable:
         self, tag_name: str, user_id: str
     ) -> int:
         with get_db() as db:
-
             return (
                 db.query(ChatIdTag)
                 .filter_by(tag_name=tag_name, user_id=user_id)
@@ -236,7 +227,6 @@ class TagTable:
     ) -> bool:
         try:
             with get_db() as db:
-
                 res = (
                     db.query(ChatIdTag)
                     .filter_by(tag_name=tag_name, chat_id=chat_id, user_id=user_id)

backend/open_webui/apps/webui/models/tools.py

@@ -1,17 +1,12 @@
-from pydantic import BaseModel, ConfigDict
-from typing import List, Optional
-import time
 import logging
-from sqlalchemy import String, Column, BigInteger, Text
+import time
+from typing import Optional
 
-from apps.webui.internal.db import Base, JSONField, get_db
-from apps.webui.models.users import Users
-
-import json
-import copy
-
-
-from config import SRC_LOG_LEVELS
+from open_webui.apps.webui.internal.db import Base, JSONField, get_db
+from open_webui.apps.webui.models.users import Users
+from open_webui.env import SRC_LOG_LEVELS
+from pydantic import BaseModel, ConfigDict
+from sqlalchemy import BigInteger, Column, String, Text
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["MODELS"])
@@ -45,7 +40,7 @@ class ToolModel(BaseModel):
     user_id: str
     name: str
     content: str
-    specs: List[dict]
+    specs: list[dict]
     meta: ToolMeta
     updated_at: int  # timestamp in epoch
     created_at: int  # timestamp in epoch
@@ -79,13 +74,10 @@ class ToolValves(BaseModel):
 
 
 class ToolsTable:
-
     def insert_new_tool(
-        self, user_id: str, form_data: ToolForm, specs: List[dict]
+        self, user_id: str, form_data: ToolForm, specs: list[dict]
     ) -> Optional[ToolModel]:
-
         with get_db() as db:
-
             tool = ToolModel(
                 **{
                     **form_data.model_dump(),
@@ -112,20 +104,18 @@ class ToolsTable:
     def get_tool_by_id(self, id: str) -> Optional[ToolModel]:
         try:
             with get_db() as db:
-
                 tool = db.get(Tool, id)
                 return ToolModel.model_validate(tool)
-        except:
+        except Exception:
             return None
 
-    def get_tools(self) -> List[ToolModel]:
+    def get_tools(self) -> list[ToolModel]:
         with get_db() as db:
             return [ToolModel.model_validate(tool) for tool in db.query(Tool).all()]
 
     def get_tool_valves_by_id(self, id: str) -> Optional[dict]:
         try:
             with get_db() as db:
-
                 tool = db.get(Tool, id)
                 return tool.valves if tool.valves else {}
         except Exception as e:
@@ -135,13 +125,12 @@ class ToolsTable:
     def update_tool_valves_by_id(self, id: str, valves: dict) -> Optional[ToolValves]:
         try:
             with get_db() as db:
-
                 db.query(Tool).filter_by(id=id).update(
                     {"valves": valves, "updated_at": int(time.time())}
                 )
                 db.commit()
                 return self.get_tool_by_id(id)
-        except:
+        except Exception:
             return None
 
     def get_user_valves_by_id_and_user_id(
@@ -196,7 +185,7 @@ class ToolsTable:
                 tool = db.query(Tool).get(id)
                 db.refresh(tool)
                 return ToolModel.model_validate(tool)
-        except:
+        except Exception:
             return None
 
     def delete_tool_by_id(self, id: str) -> bool:
@@ -206,7 +195,7 @@ class ToolsTable:
                 db.commit()
 
                 return True
-        except:
+        except Exception:
             return False
 
 

backend/open_webui/apps/webui/models/users.py

@@ -1,13 +1,10 @@
-from pydantic import BaseModel, ConfigDict, parse_obj_as
-from typing import List, Union, Optional
 import time
+from typing import Optional
 
-from sqlalchemy import String, Column, BigInteger, Text
-
-from utils.misc import get_gravatar_url
-
-from apps.webui.internal.db import Base, JSONField, Session, get_db
-from apps.webui.models.chats import Chats
+from open_webui.apps.webui.internal.db import Base, JSONField, get_db
+from open_webui.apps.webui.models.chats import Chats
+from pydantic import BaseModel, ConfigDict
+from sqlalchemy import BigInteger, Column, String, Text
 
 ####################
 # User DB Schema
@@ -78,7 +75,6 @@ class UserUpdateForm(BaseModel):
 
 
 class UsersTable:
-
     def insert_new_user(
         self,
         id: str,
@@ -116,37 +112,34 @@ class UsersTable:
             with get_db() as db:
                 user = db.query(User).filter_by(id=id).first()
                 return UserModel.model_validate(user)
-        except Exception as e:
+        except Exception:
             return None
 
     def get_user_by_api_key(self, api_key: str) -> Optional[UserModel]:
         try:
             with get_db() as db:
-
                 user = db.query(User).filter_by(api_key=api_key).first()
                 return UserModel.model_validate(user)
-        except:
+        except Exception:
             return None
 
     def get_user_by_email(self, email: str) -> Optional[UserModel]:
         try:
             with get_db() as db:
-
                 user = db.query(User).filter_by(email=email).first()
                 return UserModel.model_validate(user)
-        except:
+        except Exception:
             return None
 
     def get_user_by_oauth_sub(self, sub: str) -> Optional[UserModel]:
         try:
             with get_db() as db:
-
                 user = db.query(User).filter_by(oauth_sub=sub).first()
                 return UserModel.model_validate(user)
-        except:
+        except Exception:
             return None
 
-    def get_users(self, skip: int = 0, limit: int = 50) -> List[UserModel]:
+    def get_users(self, skip: int = 0, limit: int = 50) -> list[UserModel]:
         with get_db() as db:
             users = (
                 db.query(User)
@@ -164,7 +157,7 @@ class UsersTable:
             with get_db() as db:
                 user = db.query(User).order_by(User.created_at).first()
                 return UserModel.model_validate(user)
-        except:
+        except Exception:
             return None
 
     def update_user_role_by_id(self, id: str, role: str) -> Optional[UserModel]:
@@ -174,7 +167,7 @@ class UsersTable:
                 db.commit()
                 user = db.query(User).filter_by(id=id).first()
                 return UserModel.model_validate(user)
-        except:
+        except Exception:
             return None
 
     def update_user_profile_image_url_by_id(
@@ -189,13 +182,12 @@ class UsersTable:
 
                 user = db.query(User).filter_by(id=id).first()
                 return UserModel.model_validate(user)
-        except:
+        except Exception:
             return None
 
     def update_user_last_active_by_id(self, id: str) -> Optional[UserModel]:
         try:
             with get_db() as db:
-
                 db.query(User).filter_by(id=id).update(
                     {"last_active_at": int(time.time())}
                 )
@@ -203,7 +195,7 @@ class UsersTable:
 
                 user = db.query(User).filter_by(id=id).first()
                 return UserModel.model_validate(user)
-        except:
+        except Exception:
             return None
 
     def update_user_oauth_sub_by_id(
@@ -216,7 +208,7 @@ class UsersTable:
 
                 user = db.query(User).filter_by(id=id).first()
                 return UserModel.model_validate(user)
-        except:
+        except Exception:
             return None
 
     def update_user_by_id(self, id: str, updated: dict) -> Optional[UserModel]:
@@ -228,7 +220,7 @@ class UsersTable:
                 user = db.query(User).filter_by(id=id).first()
                 return UserModel.model_validate(user)
                 # return UserModel(**user.dict())
-        except Exception as e:
+        except Exception:
             return None
 
     def delete_user_by_id(self, id: str) -> bool:
@@ -245,7 +237,7 @@ class UsersTable:
                 return True
             else:
                 return False
-        except:
+        except Exception:
             return False
 
     def update_user_api_key_by_id(self, id: str, api_key: str) -> str:
@@ -254,7 +246,7 @@ class UsersTable:
                 result = db.query(User).filter_by(id=id).update({"api_key": api_key})
                 db.commit()
                 return True if result == 1 else False
-        except:
+        except Exception:
             return False
 
     def get_user_api_key_by_id(self, id: str) -> Optional[str]:
@@ -262,7 +254,7 @@ class UsersTable:
             with get_db() as db:
                 user = db.query(User).filter_by(id=id).first()
                 return user.api_key
-        except Exception as e:
+        except Exception:
             return None
 
 

backend/open_webui/apps/webui/routers/auths.py

@@ -1,43 +1,36 @@
-import logging
-
-from fastapi import Request, UploadFile, File
-from fastapi import Depends, HTTPException, status
-from fastapi.responses import Response
-
-from fastapi import APIRouter
-from pydantic import BaseModel
 import re
 import uuid
-import csv
 
-from apps.webui.models.auths import (
-    SigninForm,
-    SignupForm,
+from open_webui.apps.webui.models.auths import (
     AddUserForm,
-    UpdateProfileForm,
-    UpdatePasswordForm,
-    UserResponse,
-    SigninResponse,
-    Auths,
     ApiKey,
+    Auths,
+    SigninForm,
+    SigninResponse,
+    SignupForm,
+    UpdatePasswordForm,
+    UpdateProfileForm,
+    UserResponse,
 )
-from apps.webui.models.users import Users
-
-from utils.utils import (
-    get_password_hash,
-    get_current_user,
-    get_admin_user,
-    create_token,
-    create_api_key,
-)
-from utils.misc import parse_duration, validate_email_format
-from utils.webhook import post_webhook
-from constants import ERROR_MESSAGES, WEBHOOK_MESSAGES
-from config import (
-    WEBUI_AUTH,
+from open_webui.apps.webui.models.users import Users
+from open_webui.config import WEBUI_AUTH
+from open_webui.constants import ERROR_MESSAGES, WEBHOOK_MESSAGES
+from open_webui.env import (
     WEBUI_AUTH_TRUSTED_EMAIL_HEADER,
     WEBUI_AUTH_TRUSTED_NAME_HEADER,
 )
+from fastapi import APIRouter, Depends, HTTPException, Request, status
+from fastapi.responses import Response
+from pydantic import BaseModel
+from open_webui.utils.misc import parse_duration, validate_email_format
+from open_webui.utils.utils import (
+    create_api_key,
+    create_token,
+    get_admin_user,
+    get_current_user,
+    get_password_hash,
+)
+from open_webui.utils.webhook import post_webhook
 
 router = APIRouter()
 
@@ -195,7 +188,11 @@ async def signin(request: Request, response: Response, form_data: SigninForm):
 
 @router.post("/signup", response_model=SigninResponse)
 async def signup(request: Request, response: Response, form_data: SignupForm):
-    if not request.app.state.config.ENABLE_SIGNUP and WEBUI_AUTH:
+    if (
+        not request.app.state.config.ENABLE_SIGNUP
+        or not request.app.state.config.ENABLE_LOGIN_FORM
+        or not WEBUI_AUTH
+    ):
         raise HTTPException(
             status.HTTP_403_FORBIDDEN, detail=ERROR_MESSAGES.ACCESS_PROHIBITED
         )
@@ -228,7 +225,6 @@ async def signup(request: Request, response: Response, form_data: SignupForm):
                 data={"id": user.id},
                 expires_delta=parse_duration(request.app.state.config.JWT_EXPIRES_IN),
             )
-            # response.set_cookie(key='token', value=token, httponly=True)
 
             # Set the cookie token
             response.set_cookie(
@@ -270,7 +266,6 @@ async def signup(request: Request, response: Response, form_data: SignupForm):
 
 @router.post("/add", response_model=SigninResponse)
 async def add_user(form_data: AddUserForm, user=Depends(get_admin_user)):
-
     if not validate_email_format(form_data.email.lower()):
         raise HTTPException(
             status.HTTP_400_BAD_REQUEST, detail=ERROR_MESSAGES.INVALID_EMAIL_FORMAT
@@ -280,7 +275,6 @@ async def add_user(form_data: AddUserForm, user=Depends(get_admin_user)):
         raise HTTPException(400, detail=ERROR_MESSAGES.EMAIL_TAKEN)
 
     try:
-
         print(form_data)
         hashed = get_password_hash(form_data.password)
         user = Auths.insert_new_auth(
@@ -352,6 +346,7 @@ async def get_admin_config(request: Request, user=Depends(get_admin_user)):
         "DEFAULT_USER_ROLE": request.app.state.config.DEFAULT_USER_ROLE,
         "JWT_EXPIRES_IN": request.app.state.config.JWT_EXPIRES_IN,
         "ENABLE_COMMUNITY_SHARING": request.app.state.config.ENABLE_COMMUNITY_SHARING,
+        "ENABLE_MESSAGE_RATING": request.app.state.config.ENABLE_MESSAGE_RATING,
     }
 
 
@@ -361,6 +356,7 @@ class AdminConfig(BaseModel):
     DEFAULT_USER_ROLE: str
     JWT_EXPIRES_IN: str
     ENABLE_COMMUNITY_SHARING: bool
+    ENABLE_MESSAGE_RATING: bool
 
 
 @router.post("/admin/config")
@@ -382,6 +378,7 @@ async def update_admin_config(
     request.app.state.config.ENABLE_COMMUNITY_SHARING = (
         form_data.ENABLE_COMMUNITY_SHARING
     )
+    request.app.state.config.ENABLE_MESSAGE_RATING = form_data.ENABLE_MESSAGE_RATING
 
     return {
         "SHOW_ADMIN_DETAILS": request.app.state.config.SHOW_ADMIN_DETAILS,
@@ -389,6 +386,7 @@ async def update_admin_config(
         "DEFAULT_USER_ROLE": request.app.state.config.DEFAULT_USER_ROLE,
         "JWT_EXPIRES_IN": request.app.state.config.JWT_EXPIRES_IN,
         "ENABLE_COMMUNITY_SHARING": request.app.state.config.ENABLE_COMMUNITY_SHARING,
+        "ENABLE_MESSAGE_RATING": request.app.state.config.ENABLE_MESSAGE_RATING,
     }
 
 

backend/open_webui/apps/webui/routers/chats.py

@@ -1,34 +1,25 @@
-from fastapi import Depends, Request, HTTPException, status
-from datetime import datetime, timedelta
-from typing import List, Union, Optional
-from utils.utils import get_verified_user, get_admin_user
-from fastapi import APIRouter
-from pydantic import BaseModel
 import json
 import logging
+from typing import Optional
 
-from apps.webui.models.users import Users
-from apps.webui.models.chats import (
-    ChatModel,
-    ChatResponse,
-    ChatTitleForm,
+from open_webui.apps.webui.models.chats import (
     ChatForm,
-    ChatTitleIdResponse,
+    ChatResponse,
     Chats,
+    ChatTitleIdResponse,
 )
-
-
-from apps.webui.models.tags import (
-    TagModel,
-    ChatIdTagModel,
+from open_webui.apps.webui.models.tags import (
     ChatIdTagForm,
-    ChatTagsResponse,
+    ChatIdTagModel,
+    TagModel,
     Tags,
 )
-
-from constants import ERROR_MESSAGES
-
-from config import SRC_LOG_LEVELS, ENABLE_ADMIN_EXPORT, ENABLE_ADMIN_CHAT_ACCESS
+from open_webui.config import ENABLE_ADMIN_CHAT_ACCESS, ENABLE_ADMIN_EXPORT
+from open_webui.constants import ERROR_MESSAGES
+from open_webui.env import SRC_LOG_LEVELS
+from fastapi import APIRouter, Depends, HTTPException, Request, status
+from pydantic import BaseModel
+from open_webui.utils.utils import get_admin_user, get_verified_user
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["MODELS"])
@@ -40,8 +31,8 @@ router = APIRouter()
 ############################
 
 
-@router.get("/", response_model=List[ChatTitleIdResponse])
-@router.get("/list", response_model=List[ChatTitleIdResponse])
+@router.get("/", response_model=list[ChatTitleIdResponse])
+@router.get("/list", response_model=list[ChatTitleIdResponse])
 async def get_session_user_chat_list(
     user=Depends(get_verified_user), page: Optional[int] = None
 ):
@@ -61,7 +52,6 @@ async def get_session_user_chat_list(
 
 @router.delete("/", response_model=bool)
 async def delete_all_user_chats(request: Request, user=Depends(get_verified_user)):
-
     if (
         user.role == "user"
         and not request.app.state.config.USER_PERMISSIONS["chat"]["deletion"]
@@ -80,7 +70,7 @@ async def delete_all_user_chats(request: Request, user=Depends(get_verified_user
 ############################
 
 
-@router.get("/list/user/{user_id}", response_model=List[ChatTitleIdResponse])
+@router.get("/list/user/{user_id}", response_model=list[ChatTitleIdResponse])
 async def get_user_chat_list_by_user_id(
     user_id: str,
     user=Depends(get_admin_user),
@@ -119,7 +109,7 @@ async def create_new_chat(form_data: ChatForm, user=Depends(get_verified_user)):
 ############################
 
 
-@router.get("/all", response_model=List[ChatResponse])
+@router.get("/all", response_model=list[ChatResponse])
 async def get_user_chats(user=Depends(get_verified_user)):
     return [
         ChatResponse(**{**chat.model_dump(), "chat": json.loads(chat.chat)})
@@ -132,7 +122,7 @@ async def get_user_chats(user=Depends(get_verified_user)):
 ############################
 
 
-@router.get("/all/archived", response_model=List[ChatResponse])
+@router.get("/all/archived", response_model=list[ChatResponse])
 async def get_user_archived_chats(user=Depends(get_verified_user)):
     return [
         ChatResponse(**{**chat.model_dump(), "chat": json.loads(chat.chat)})
@@ -145,7 +135,7 @@ async def get_user_archived_chats(user=Depends(get_verified_user)):
 ############################
 
 
-@router.get("/all/db", response_model=List[ChatResponse])
+@router.get("/all/db", response_model=list[ChatResponse])
 async def get_all_user_chats_in_db(user=Depends(get_admin_user)):
     if not ENABLE_ADMIN_EXPORT:
         raise HTTPException(
@@ -163,7 +153,7 @@ async def get_all_user_chats_in_db(user=Depends(get_admin_user)):
 ############################
 
 
-@router.get("/archived", response_model=List[ChatTitleIdResponse])
+@router.get("/archived", response_model=list[ChatTitleIdResponse])
 async def get_archived_session_user_chat_list(
     user=Depends(get_verified_user), skip: int = 0, limit: int = 50
 ):
@@ -216,11 +206,10 @@ class TagNameForm(BaseModel):
     limit: Optional[int] = 50
 
 
-@router.post("/tags", response_model=List[ChatTitleIdResponse])
+@router.post("/tags", response_model=list[ChatTitleIdResponse])
 async def get_user_chat_list_by_tag_name(
     form_data: TagNameForm, user=Depends(get_verified_user)
 ):
-
     chat_ids = [
         chat_id_tag.chat_id
         for chat_id_tag in Tags.get_chat_ids_by_tag_name_and_user_id(
@@ -241,7 +230,7 @@ async def get_user_chat_list_by_tag_name(
 ############################
 
 
-@router.get("/tags/all", response_model=List[TagModel])
+@router.get("/tags/all", response_model=list[TagModel])
 async def get_all_tags(user=Depends(get_verified_user)):
     try:
         tags = Tags.get_tags_by_user_id(user.id)
@@ -299,7 +288,6 @@ async def update_chat_by_id(
 
 @router.delete("/{id}", response_model=bool)
 async def delete_chat_by_id(request: Request, id: str, user=Depends(get_verified_user)):
-
     if user.role == "admin":
         result = Chats.delete_chat_by_id(id)
         return result
@@ -323,7 +311,6 @@ async def delete_chat_by_id(request: Request, id: str, user=Depends(get_verified
 async def clone_chat_by_id(id: str, user=Depends(get_verified_user)):
     chat = Chats.get_chat_by_id_and_user_id(id, user.id)
     if chat:
-
         chat_body = json.loads(chat.chat)
         updated_chat = {
             **chat_body,
@@ -417,7 +404,7 @@ async def delete_shared_chat_by_id(id: str, user=Depends(get_verified_user)):
 ############################
 
 
-@router.get("/{id}/tags", response_model=List[TagModel])
+@router.get("/{id}/tags", response_model=list[TagModel])
 async def get_chat_tags_by_id(id: str, user=Depends(get_verified_user)):
     tags = Tags.get_tags_by_chat_id_and_user_id(id, user.id)
 

backend/open_webui/apps/webui/routers/configs.py

@@ -1,40 +1,50 @@
-from fastapi import Response, Request
-from fastapi import Depends, FastAPI, HTTPException, status
-from datetime import datetime, timedelta
-from typing import List, Union
-
-from fastapi import APIRouter
+from open_webui.config import BannerModel
+from fastapi import APIRouter, Depends, Request
 from pydantic import BaseModel
-import time
-import uuid
+from open_webui.utils.utils import get_admin_user, get_verified_user
 
-from config import BannerModel
 
-from apps.webui.models.users import Users
-
-from utils.utils import (
-    get_password_hash,
-    get_verified_user,
-    get_admin_user,
-    create_token,
-)
-from utils.misc import get_gravatar_url, validate_email_format
-from constants import ERROR_MESSAGES
+from open_webui.config import get_config, save_config
 
 router = APIRouter()
 
 
+############################
+# ImportConfig
+############################
+
+
+class ImportConfigForm(BaseModel):
+    config: dict
+
+
+@router.post("/import", response_model=dict)
+async def import_config(form_data: ImportConfigForm, user=Depends(get_admin_user)):
+    save_config(form_data.config)
+    return get_config()
+
+
+############################
+# ExportConfig
+############################
+
+
+@router.get("/export", response_model=dict)
+async def export_config(user=Depends(get_admin_user)):
+    return get_config()
+
+
 class SetDefaultModelsForm(BaseModel):
     models: str
 
 
 class PromptSuggestion(BaseModel):
-    title: List[str]
+    title: list[str]
     content: str
 
 
 class SetDefaultSuggestionsForm(BaseModel):
-    suggestions: List[PromptSuggestion]
+    suggestions: list[PromptSuggestion]
 
 
 ############################
@@ -50,7 +60,7 @@ async def set_global_default_models(
     return request.app.state.config.DEFAULT_MODELS
 
 
-@router.post("/default/suggestions", response_model=List[PromptSuggestion])
+@router.post("/default/suggestions", response_model=list[PromptSuggestion])
 async def set_global_default_suggestions(
     request: Request,
     form_data: SetDefaultSuggestionsForm,
@@ -67,10 +77,10 @@ async def set_global_default_suggestions(
 
 
 class SetBannersForm(BaseModel):
-    banners: List[BannerModel]
+    banners: list[BannerModel]
 
 
-@router.post("/banners", response_model=List[BannerModel])
+@router.post("/banners", response_model=list[BannerModel])
 async def set_banners(
     request: Request,
     form_data: SetBannersForm,
@@ -81,7 +91,7 @@ async def set_banners(
     return request.app.state.config.BANNERS
 
 
-@router.get("/banners", response_model=List[BannerModel])
+@router.get("/banners", response_model=list[BannerModel])
 async def get_banners(
     request: Request,
     user=Depends(get_verified_user),

backend/open_webui/apps/webui/routers/documents.py

@@ -1,21 +1,16 @@
-from fastapi import Depends, FastAPI, HTTPException, status
-from datetime import datetime, timedelta
-from typing import List, Union, Optional
-
-from fastapi import APIRouter
-from pydantic import BaseModel
 import json
+from typing import Optional
 
-from apps.webui.models.documents import (
-    Documents,
+from open_webui.apps.webui.models.documents import (
     DocumentForm,
-    DocumentUpdateForm,
-    DocumentModel,
     DocumentResponse,
+    Documents,
+    DocumentUpdateForm,
 )
-
-from utils.utils import get_verified_user, get_admin_user
-from constants import ERROR_MESSAGES
+from open_webui.constants import ERROR_MESSAGES
+from fastapi import APIRouter, Depends, HTTPException, status
+from pydantic import BaseModel
+from open_webui.utils.utils import get_admin_user, get_verified_user
 
 router = APIRouter()
 
@@ -24,7 +19,7 @@ router = APIRouter()
 ############################
 
 
-@router.get("/", response_model=List[DocumentResponse])
+@router.get("/", response_model=list[DocumentResponse])
 async def get_documents(user=Depends(get_verified_user)):
     docs = [
         DocumentResponse(
@@ -46,7 +41,7 @@ async def get_documents(user=Depends(get_verified_user)):
 @router.post("/create", response_model=Optional[DocumentResponse])
 async def create_new_doc(form_data: DocumentForm, user=Depends(get_admin_user)):
     doc = Documents.get_doc_by_name(form_data.name)
-    if doc == None:
+    if doc is None:
         doc = Documents.insert_new_doc(user.id, form_data)
 
         if doc:
@@ -102,7 +97,7 @@ class TagItem(BaseModel):
 
 class TagDocumentForm(BaseModel):
     name: str
-    tags: List[dict]
+    tags: list[dict]
 
 
 @router.post("/doc/tags", response_model=Optional[DocumentResponse])

backend/open_webui/apps/webui/routers/files.py

@@ -1,42 +1,17 @@
-from fastapi import (
-    Depends,
-    FastAPI,
-    HTTPException,
-    status,
-    Request,
-    UploadFile,
-    File,
-    Form,
-)
-
-
-from datetime import datetime, timedelta
-from typing import List, Union, Optional
-from pathlib import Path
-
-from fastapi import APIRouter
-from fastapi.responses import StreamingResponse, JSONResponse, FileResponse
-
-from pydantic import BaseModel
-import json
-
-from apps.webui.models.files import (
-    Files,
-    FileForm,
-    FileModel,
-    FileModelResponse,
-)
-from utils.utils import get_verified_user, get_admin_user
-from constants import ERROR_MESSAGES
-
-from importlib import util
+import logging
 import os
+import shutil
 import uuid
-import os, shutil, logging, re
-
-
-from config import SRC_LOG_LEVELS, UPLOAD_DIR
+from pathlib import Path
+from typing import Optional
 
+from open_webui.apps.webui.models.files import FileForm, FileModel, Files
+from open_webui.config import UPLOAD_DIR
+from open_webui.constants import ERROR_MESSAGES
+from open_webui.env import SRC_LOG_LEVELS
+from fastapi import APIRouter, Depends, File, HTTPException, UploadFile, status
+from fastapi.responses import FileResponse
+from open_webui.utils.utils import get_admin_user, get_verified_user
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["MODELS"])
@@ -104,9 +79,12 @@ def upload_file(file: UploadFile = File(...), user=Depends(get_verified_user)):
 ############################
 
 
-@router.get("/", response_model=List[FileModel])
+@router.get("/", response_model=list[FileModel])
 async def list_files(user=Depends(get_verified_user)):
-    files = Files.get_files()
+    if user.role == "admin":
+        files = Files.get_files()
+    else:
+        files = Files.get_files_by_user_id(user.id)
     return files
 
 
@@ -156,7 +134,7 @@ async def delete_all_files(user=Depends(get_admin_user)):
 async def get_file_by_id(id: str, user=Depends(get_verified_user)):
     file = Files.get_file_by_id(id)
 
-    if file:
+    if file and (file.user_id == user.id or user.role == "admin"):
         return file
     else:
         raise HTTPException(
@@ -174,7 +152,7 @@ async def get_file_by_id(id: str, user=Depends(get_verified_user)):
 async def get_file_content_by_id(id: str, user=Depends(get_verified_user)):
     file = Files.get_file_by_id(id)
 
-    if file:
+    if file and (file.user_id == user.id or user.role == "admin"):
         file_path = Path(file.meta["path"])
 
         # Check if the file already exists in the cache
@@ -197,7 +175,7 @@ async def get_file_content_by_id(id: str, user=Depends(get_verified_user)):
 async def get_file_content_by_id(id: str, user=Depends(get_verified_user)):
     file = Files.get_file_by_id(id)
 
-    if file:
+    if file and (file.user_id == user.id or user.role == "admin"):
         file_path = Path(file.meta["path"])
 
         # Check if the file already exists in the cache
@@ -224,8 +202,7 @@ async def get_file_content_by_id(id: str, user=Depends(get_verified_user)):
 @router.delete("/{id}")
 async def delete_file_by_id(id: str, user=Depends(get_verified_user)):
     file = Files.get_file_by_id(id)
-
-    if file:
+    if file and (file.user_id == user.id or user.role == "admin"):
         result = Files.delete_file_by_id(id)
         if result:
             return {"message": "File deleted successfully"}

backend/open_webui/apps/webui/routers/functions.py

@@ -1,27 +1,18 @@
-from fastapi import Depends, FastAPI, HTTPException, status, Request
-from datetime import datetime, timedelta
-from typing import List, Union, Optional
+import os
+from pathlib import Path
+from typing import Optional
 
-from fastapi import APIRouter
-from pydantic import BaseModel
-import json
-
-from apps.webui.models.functions import (
-    Functions,
+from open_webui.apps.webui.models.functions import (
     FunctionForm,
     FunctionModel,
     FunctionResponse,
+    Functions,
 )
-from apps.webui.utils import load_function_module_by_id
-from utils.utils import get_verified_user, get_admin_user
-from constants import ERROR_MESSAGES
-
-from importlib import util
-import os
-from pathlib import Path
-
-from config import DATA_DIR, CACHE_DIR, FUNCTIONS_DIR
-
+from open_webui.apps.webui.utils import load_function_module_by_id, replace_imports
+from open_webui.config import CACHE_DIR, FUNCTIONS_DIR
+from open_webui.constants import ERROR_MESSAGES
+from fastapi import APIRouter, Depends, HTTPException, Request, status
+from open_webui.utils.utils import get_admin_user, get_verified_user
 
 router = APIRouter()
 
@@ -30,7 +21,7 @@ router = APIRouter()
 ############################
 
 
-@router.get("/", response_model=List[FunctionResponse])
+@router.get("/", response_model=list[FunctionResponse])
 async def get_functions(user=Depends(get_verified_user)):
     return Functions.get_functions()
 
@@ -40,7 +31,7 @@ async def get_functions(user=Depends(get_verified_user)):
 ############################
 
 
-@router.get("/export", response_model=List[FunctionModel])
+@router.get("/export", response_model=list[FunctionModel])
 async def get_functions(user=Depends(get_admin_user)):
     return Functions.get_functions()
 
@@ -63,14 +54,12 @@ async def create_new_function(
     form_data.id = form_data.id.lower()
 
     function = Functions.get_function_by_id(form_data.id)
-    if function == None:
-        function_path = os.path.join(FUNCTIONS_DIR, f"{form_data.id}.py")
+    if function is None:
         try:
-            with open(function_path, "w") as function_file:
-                function_file.write(form_data.content)
-
+            form_data.content = replace_imports(form_data.content)
             function_module, function_type, frontmatter = load_function_module_by_id(
-                form_data.id
+                form_data.id,
+                content=form_data.content,
             )
             form_data.meta.manifest = frontmatter
 
@@ -183,13 +172,11 @@ async def toggle_global_by_id(id: str, user=Depends(get_admin_user)):
 async def update_function_by_id(
     request: Request, id: str, form_data: FunctionForm, user=Depends(get_admin_user)
 ):
-    function_path = os.path.join(FUNCTIONS_DIR, f"{id}.py")
-
     try:
-        with open(function_path, "w") as function_file:
-            function_file.write(form_data.content)
-
-        function_module, function_type, frontmatter = load_function_module_by_id(id)
+        form_data.content = replace_imports(form_data.content)
+        function_module, function_type, frontmatter = load_function_module_by_id(
+            id, content=form_data.content
+        )
         form_data.meta.manifest = frontmatter
 
         FUNCTIONS = request.app.state.FUNCTIONS
@@ -231,13 +218,6 @@ async def delete_function_by_id(
         if id in FUNCTIONS:
             del FUNCTIONS[id]
 
-        # delete the function file
-        function_path = os.path.join(FUNCTIONS_DIR, f"{id}.py")
-        try:
-            os.remove(function_path)
-        except:
-            pass
-
     return result
 
 
@@ -304,7 +284,6 @@ async def update_function_valves_by_id(
 ):
     function = Functions.get_function_by_id(id)
     if function:
-
         if id in request.app.state.FUNCTIONS:
             function_module = request.app.state.FUNCTIONS[id]
         else:

backend/open_webui/apps/webui/routers/memories.py

@@ -1,18 +1,13 @@
-from fastapi import Response, Request
-from fastapi import Depends, FastAPI, HTTPException, status
-from datetime import datetime, timedelta
-from typing import List, Union, Optional
-
-from fastapi import APIRouter
+from fastapi import APIRouter, Depends, HTTPException, Request
 from pydantic import BaseModel
 import logging
+from typing import Optional
 
-from apps.webui.models.memories import Memories, MemoryModel
+from open_webui.apps.webui.models.memories import Memories, MemoryModel
+from open_webui.apps.rag.vector.connector import VECTOR_DB_CLIENT
+from open_webui.utils.utils import get_verified_user
+from open_webui.env import SRC_LOG_LEVELS
 
-from utils.utils import get_verified_user
-from constants import ERROR_MESSAGES
-
-from config import SRC_LOG_LEVELS, CHROMA_CLIENT
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["MODELS"])
@@ -30,7 +25,7 @@ async def get_embeddings(request: Request):
 ############################
 
 
-@router.get("/", response_model=List[MemoryModel])
+@router.get("/", response_model=list[MemoryModel])
 async def get_memories(user=Depends(get_verified_user)):
     return Memories.get_memories_by_user_id(user.id)
 
@@ -55,47 +50,22 @@ async def add_memory(
     user=Depends(get_verified_user),
 ):
     memory = Memories.insert_new_memory(user.id, form_data.content)
-    memory_embedding = request.app.state.EMBEDDING_FUNCTION(memory.content)
 
-    collection = CHROMA_CLIENT.get_or_create_collection(name=f"user-memory-{user.id}")
-    collection.upsert(
-        documents=[memory.content],
-        ids=[memory.id],
-        embeddings=[memory_embedding],
-        metadatas=[{"created_at": memory.created_at}],
+    VECTOR_DB_CLIENT.upsert(
+        collection_name=f"user-memory-{user.id}",
+        items=[
+            {
+                "id": memory.id,
+                "text": memory.content,
+                "vector": request.app.state.EMBEDDING_FUNCTION(memory.content),
+                "metadata": {"created_at": memory.created_at},
+            }
+        ],
     )
 
     return memory
 
 
-@router.post("/{memory_id}/update", response_model=Optional[MemoryModel])
-async def update_memory_by_id(
-    memory_id: str,
-    request: Request,
-    form_data: MemoryUpdateModel,
-    user=Depends(get_verified_user),
-):
-    memory = Memories.update_memory_by_id(memory_id, form_data.content)
-    if memory is None:
-        raise HTTPException(status_code=404, detail="Memory not found")
-
-    if form_data.content is not None:
-        memory_embedding = request.app.state.EMBEDDING_FUNCTION(form_data.content)
-        collection = CHROMA_CLIENT.get_or_create_collection(
-            name=f"user-memory-{user.id}"
-        )
-        collection.upsert(
-            documents=[form_data.content],
-            ids=[memory.id],
-            embeddings=[memory_embedding],
-            metadatas=[
-                {"created_at": memory.created_at, "updated_at": memory.updated_at}
-            ],
-        )
-
-    return memory
-
-
 ############################
 # QueryMemory
 ############################
@@ -110,12 +80,10 @@ class QueryMemoryForm(BaseModel):
 async def query_memory(
     request: Request, form_data: QueryMemoryForm, user=Depends(get_verified_user)
 ):
-    query_embedding = request.app.state.EMBEDDING_FUNCTION(form_data.content)
-    collection = CHROMA_CLIENT.get_or_create_collection(name=f"user-memory-{user.id}")
-
-    results = collection.query(
-        query_embeddings=[query_embedding],
-        n_results=form_data.k,  # how many results to return
+    results = VECTOR_DB_CLIENT.search(
+        collection_name=f"user-memory-{user.id}",
+        vectors=[request.app.state.EMBEDDING_FUNCTION(form_data.content)],
+        limit=form_data.k,
     )
 
     return results
@@ -124,21 +92,29 @@ async def query_memory(
 ############################
 # ResetMemoryFromVectorDB
 ############################
-@router.get("/reset", response_model=bool)
+@router.post("/reset", response_model=bool)
 async def reset_memory_from_vector_db(
     request: Request, user=Depends(get_verified_user)
 ):
-    CHROMA_CLIENT.delete_collection(f"user-memory-{user.id}")
-    collection = CHROMA_CLIENT.get_or_create_collection(name=f"user-memory-{user.id}")
+    VECTOR_DB_CLIENT.delete_collection(f"user-memory-{user.id}")
 
     memories = Memories.get_memories_by_user_id(user.id)
-    for memory in memories:
-        memory_embedding = request.app.state.EMBEDDING_FUNCTION(memory.content)
-        collection.upsert(
-            documents=[memory.content],
-            ids=[memory.id],
-            embeddings=[memory_embedding],
-        )
+    VECTOR_DB_CLIENT.upsert(
+        collection_name=f"user-memory-{user.id}",
+        items=[
+            {
+                "id": memory.id,
+                "text": memory.content,
+                "vector": request.app.state.EMBEDDING_FUNCTION(memory.content),
+                "metadata": {
+                    "created_at": memory.created_at,
+                    "updated_at": memory.updated_at,
+                },
+            }
+            for memory in memories
+        ],
+    )
+
     return True
 
 
@@ -147,13 +123,13 @@ async def reset_memory_from_vector_db(
 ############################
 
 
-@router.delete("/user", response_model=bool)
+@router.delete("/delete/user", response_model=bool)
 async def delete_memory_by_user_id(user=Depends(get_verified_user)):
     result = Memories.delete_memories_by_user_id(user.id)
 
     if result:
         try:
-            CHROMA_CLIENT.delete_collection(f"user-memory-{user.id}")
+            VECTOR_DB_CLIENT.delete_collection(f"user-memory-{user.id}")
         except Exception as e:
             log.error(e)
         return True
@@ -161,6 +137,41 @@ async def delete_memory_by_user_id(user=Depends(get_verified_user)):
     return False
 
 
+############################
+# UpdateMemoryById
+############################
+
+
+@router.post("/{memory_id}/update", response_model=Optional[MemoryModel])
+async def update_memory_by_id(
+    memory_id: str,
+    request: Request,
+    form_data: MemoryUpdateModel,
+    user=Depends(get_verified_user),
+):
+    memory = Memories.update_memory_by_id(memory_id, form_data.content)
+    if memory is None:
+        raise HTTPException(status_code=404, detail="Memory not found")
+
+    if form_data.content is not None:
+        VECTOR_DB_CLIENT.upsert(
+            collection_name=f"user-memory-{user.id}",
+            items=[
+                {
+                    "id": memory.id,
+                    "text": memory.content,
+                    "vector": request.app.state.EMBEDDING_FUNCTION(memory.content),
+                    "metadata": {
+                        "created_at": memory.created_at,
+                        "updated_at": memory.updated_at,
+                    },
+                }
+            ],
+        )
+
+    return memory
+
+
 ############################
 # DeleteMemoryById
 ############################
@@ -171,10 +182,9 @@ async def delete_memory_by_id(memory_id: str, user=Depends(get_verified_user)):
     result = Memories.delete_memory_by_id_and_user_id(memory_id, user.id)
 
     if result:
-        collection = CHROMA_CLIENT.get_or_create_collection(
-            name=f"user-memory-{user.id}"
+        VECTOR_DB_CLIENT.delete(
+            collection_name=f"user-memory-{user.id}", ids=[memory_id]
         )
-        collection.delete(ids=[memory_id])
         return True
 
     return False

backend/open_webui/apps/webui/routers/models.py

@@ -1,15 +1,14 @@
-from fastapi import Depends, FastAPI, HTTPException, status, Request
-from datetime import datetime, timedelta
-from typing import List, Union, Optional
+from typing import Optional
 
-from fastapi import APIRouter
-from pydantic import BaseModel
-import json
-
-from apps.webui.models.models import Models, ModelModel, ModelForm, ModelResponse
-
-from utils.utils import get_verified_user, get_admin_user
-from constants import ERROR_MESSAGES
+from open_webui.apps.webui.models.models import (
+    ModelForm,
+    ModelModel,
+    ModelResponse,
+    Models,
+)
+from open_webui.constants import ERROR_MESSAGES
+from fastapi import APIRouter, Depends, HTTPException, Request, status
+from open_webui.utils.utils import get_admin_user, get_verified_user
 
 router = APIRouter()
 
@@ -18,9 +17,19 @@ router = APIRouter()
 ###########################
 
 
-@router.get("/", response_model=List[ModelResponse])
-async def get_models(user=Depends(get_verified_user)):
-    return Models.get_all_models()
+@router.get("/", response_model=list[ModelResponse])
+async def get_models(id: Optional[str] = None, user=Depends(get_verified_user)):
+    if id:
+        model = Models.get_model_by_id(id)
+        if model:
+            return [model]
+        else:
+            raise HTTPException(
+                status_code=status.HTTP_401_UNAUTHORIZED,
+                detail=ERROR_MESSAGES.NOT_FOUND,
+            )
+    else:
+        return Models.get_all_models()
 
 
 ############################
@@ -51,24 +60,6 @@ async def add_new_model(
             )
 
 
-############################
-# GetModelById
-############################
-
-
-@router.get("/", response_model=Optional[ModelModel])
-async def get_model_by_id(id: str, user=Depends(get_verified_user)):
-    model = Models.get_model_by_id(id)
-
-    if model:
-        return model
-    else:
-        raise HTTPException(
-            status_code=status.HTTP_401_UNAUTHORIZED,
-            detail=ERROR_MESSAGES.NOT_FOUND,
-        )
-
-
 ############################
 # UpdateModelById
 ############################

backend/open_webui/apps/webui/routers/prompts.py

@@ -1,15 +1,9 @@
-from fastapi import Depends, FastAPI, HTTPException, status
-from datetime import datetime, timedelta
-from typing import List, Union, Optional
+from typing import Optional
 
-from fastapi import APIRouter
-from pydantic import BaseModel
-import json
-
-from apps.webui.models.prompts import Prompts, PromptForm, PromptModel
-
-from utils.utils import get_verified_user, get_admin_user
-from constants import ERROR_MESSAGES
+from open_webui.apps.webui.models.prompts import PromptForm, PromptModel, Prompts
+from open_webui.constants import ERROR_MESSAGES
+from fastapi import APIRouter, Depends, HTTPException, status
+from open_webui.utils.utils import get_admin_user, get_verified_user
 
 router = APIRouter()
 
@@ -18,7 +12,7 @@ router = APIRouter()
 ############################
 
 
-@router.get("/", response_model=List[PromptModel])
+@router.get("/", response_model=list[PromptModel])
 async def get_prompts(user=Depends(get_verified_user)):
     return Prompts.get_prompts()
 
@@ -31,7 +25,7 @@ async def get_prompts(user=Depends(get_verified_user)):
 @router.post("/create", response_model=Optional[PromptModel])
 async def create_new_prompt(form_data: PromptForm, user=Depends(get_admin_user)):
     prompt = Prompts.get_prompt_by_command(form_data.command)
-    if prompt == None:
+    if prompt is None:
         prompt = Prompts.insert_new_prompt(user.id, form_data)
 
         if prompt:

backend/open_webui/apps/webui/routers/tools.py

@@ -1,20 +1,14 @@
-from fastapi import Depends, HTTPException, status, Request
-from typing import List, Optional
-
-from fastapi import APIRouter
-
-from apps.webui.models.tools import Tools, ToolForm, ToolModel, ToolResponse
-from apps.webui.utils import load_toolkit_module_by_id
-
-from utils.utils import get_admin_user, get_verified_user
-from utils.tools import get_tools_specs
-from constants import ERROR_MESSAGES
-
 import os
 from pathlib import Path
+from typing import Optional
 
-from config import DATA_DIR, CACHE_DIR
-
+from open_webui.apps.webui.models.tools import ToolForm, ToolModel, ToolResponse, Tools
+from open_webui.apps.webui.utils import load_toolkit_module_by_id, replace_imports
+from open_webui.config import CACHE_DIR, DATA_DIR
+from open_webui.constants import ERROR_MESSAGES
+from fastapi import APIRouter, Depends, HTTPException, Request, status
+from open_webui.utils.tools import get_tools_specs
+from open_webui.utils.utils import get_admin_user, get_verified_user
 
 TOOLS_DIR = f"{DATA_DIR}/tools"
 os.makedirs(TOOLS_DIR, exist_ok=True)
@@ -27,7 +21,7 @@ router = APIRouter()
 ############################
 
 
-@router.get("/", response_model=List[ToolResponse])
+@router.get("/", response_model=list[ToolResponse])
 async def get_toolkits(user=Depends(get_verified_user)):
     toolkits = [toolkit for toolkit in Tools.get_tools()]
     return toolkits
@@ -38,7 +32,7 @@ async def get_toolkits(user=Depends(get_verified_user)):
 ############################
 
 
-@router.get("/export", response_model=List[ToolModel])
+@router.get("/export", response_model=list[ToolModel])
 async def get_toolkits(user=Depends(get_admin_user)):
     toolkits = [toolkit for toolkit in Tools.get_tools()]
     return toolkits
@@ -65,12 +59,11 @@ async def create_new_toolkit(
 
     toolkit = Tools.get_tool_by_id(form_data.id)
     if toolkit is None:
-        toolkit_path = os.path.join(TOOLS_DIR, f"{form_data.id}.py")
         try:
-            with open(toolkit_path, "w") as tool_file:
-                tool_file.write(form_data.content)
-
-            toolkit_module, frontmatter = load_toolkit_module_by_id(form_data.id)
+            form_data.content = replace_imports(form_data.content)
+            toolkit_module, frontmatter = load_toolkit_module_by_id(
+                form_data.id, content=form_data.content
+            )
             form_data.meta.manifest = frontmatter
 
             TOOLS = request.app.state.TOOLS
@@ -132,13 +125,11 @@ async def update_toolkit_by_id(
     form_data: ToolForm,
     user=Depends(get_admin_user),
 ):
-    toolkit_path = os.path.join(TOOLS_DIR, f"{id}.py")
-
     try:
-        with open(toolkit_path, "w") as tool_file:
-            tool_file.write(form_data.content)
-
-        toolkit_module, frontmatter = load_toolkit_module_by_id(id)
+        form_data.content = replace_imports(form_data.content)
+        toolkit_module, frontmatter = load_toolkit_module_by_id(
+            id, content=form_data.content
+        )
         form_data.meta.manifest = frontmatter
 
         TOOLS = request.app.state.TOOLS
@@ -183,10 +174,6 @@ async def delete_toolkit_by_id(request: Request, id: str, user=Depends(get_admin
         if id in TOOLS:
             del TOOLS[id]
 
-        # delete the toolkit file
-        toolkit_path = os.path.join(TOOLS_DIR, f"{id}.py")
-        os.remove(toolkit_path)
-
     return result
 
 

backend/open_webui/apps/webui/routers/users.py

@@ -1,33 +1,20 @@
-from fastapi import Response, Request
-from fastapi import Depends, FastAPI, HTTPException, status
-from datetime import datetime, timedelta
-from typing import List, Union, Optional
-
-from fastapi import APIRouter
-from pydantic import BaseModel
-import time
-import uuid
 import logging
+from typing import Optional
 
-from apps.webui.models.users import (
+from open_webui.apps.webui.models.auths import Auths
+from open_webui.apps.webui.models.chats import Chats
+from open_webui.apps.webui.models.users import (
     UserModel,
-    UserUpdateForm,
     UserRoleUpdateForm,
-    UserSettings,
     Users,
+    UserSettings,
+    UserUpdateForm,
 )
-from apps.webui.models.auths import Auths
-from apps.webui.models.chats import Chats
-
-from utils.utils import (
-    get_verified_user,
-    get_password_hash,
-    get_current_user,
-    get_admin_user,
-)
-from constants import ERROR_MESSAGES
-
-from config import SRC_LOG_LEVELS
+from open_webui.constants import ERROR_MESSAGES
+from open_webui.env import SRC_LOG_LEVELS
+from fastapi import APIRouter, Depends, HTTPException, Request, status
+from pydantic import BaseModel
+from open_webui.utils.utils import get_admin_user, get_password_hash, get_verified_user
 
 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["MODELS"])
@@ -39,7 +26,7 @@ router = APIRouter()
 ############################
 
 
-@router.get("/", response_model=List[UserModel])
+@router.get("/", response_model=list[UserModel])
 async def get_users(skip: int = 0, limit: int = 50, user=Depends(get_admin_user)):
     return Users.get_users(skip, limit)
 
@@ -69,7 +56,6 @@ async def update_user_permissions(
 
 @router.post("/update/role", response_model=Optional[UserModel])
 async def update_user_role(form_data: UserRoleUpdateForm, user=Depends(get_admin_user)):
-
     if user.id != form_data.id and form_data.id != Users.get_first_user().id:
         return Users.update_user_role_by_id(form_data.id, form_data.role)
 
@@ -173,7 +159,6 @@ class UserResponse(BaseModel):
 
 @router.get("/{user_id}", response_model=UserResponse)
 async def get_user_by_id(user_id: str, user=Depends(get_verified_user)):
-
     # Check if user_id is a shared chat
     # If it is, get the user_id from the chat
     if user_id.startswith("shared-"):

backend/open_webui/apps/webui/routers/utils.py

@@ -1,23 +1,17 @@
-from pathlib import Path
 import site
+from pathlib import Path
 
-from fastapi import APIRouter, UploadFile, File, Response
-from fastapi import Depends, HTTPException, status
-from starlette.responses import StreamingResponse, FileResponse
-from pydantic import BaseModel
-
-
-from fpdf import FPDF
-import markdown
 import black
-
-
-from utils.utils import get_admin_user
-from utils.misc import calculate_sha256, get_gravatar_url
-
-from config import OLLAMA_BASE_URLS, DATA_DIR, UPLOAD_DIR, ENABLE_ADMIN_EXPORT
-from constants import ERROR_MESSAGES
-from typing import List
+import markdown
+from open_webui.config import DATA_DIR, ENABLE_ADMIN_EXPORT
+from open_webui.env import FONTS_DIR
+from open_webui.constants import ERROR_MESSAGES
+from fastapi import APIRouter, Depends, HTTPException, Response, status
+from fpdf import FPDF
+from pydantic import BaseModel
+from starlette.responses import FileResponse
+from open_webui.utils.misc import get_gravatar_url
+from open_webui.utils.utils import get_admin_user
 
 router = APIRouter()
 
@@ -57,21 +51,18 @@ async def get_html_from_markdown(
 
 class ChatForm(BaseModel):
     title: str
-    messages: List[dict]
+    messages: list[dict]
 
 
 @router.post("/pdf")
 async def download_chat_as_pdf(
     form_data: ChatForm,
 ):
+    global FONTS_DIR
+
     pdf = FPDF()
     pdf.add_page()
 
-    # When running in docker, workdir is /app/backend, so fonts is in /app/backend/static/fonts
-    FONTS_DIR = Path("./static/fonts")
-
-    # Non Docker Installation
-
     # When running using `pip install` the static directory is in the site packages.
     if not FONTS_DIR.exists():
         FONTS_DIR = Path(site.getsitepackages()[0]) / "static/fonts"
@@ -85,9 +76,10 @@ async def download_chat_as_pdf(
     pdf.add_font("NotoSans", "i", f"{FONTS_DIR}/NotoSans-Italic.ttf")
     pdf.add_font("NotoSansKR", "", f"{FONTS_DIR}/NotoSansKR-Regular.ttf")
     pdf.add_font("NotoSansJP", "", f"{FONTS_DIR}/NotoSansJP-Regular.ttf")
+    pdf.add_font("NotoSansSC", "", f"{FONTS_DIR}/NotoSansSC-Regular.ttf")
 
     pdf.set_font("NotoSans", size=12)
-    pdf.set_fallback_fonts(["NotoSansKR", "NotoSansJP"])
+    pdf.set_fallback_fonts(["NotoSansKR", "NotoSansJP", "NotoSansSC"])
 
     pdf.set_auto_page_break(auto=True, margin=15)
 
@@ -114,7 +106,7 @@ async def download_chat_as_pdf(
     return Response(
         content=bytes(pdf_bytes),
         media_type="application/pdf",
-        headers={"Content-Disposition": f"attachment;filename=chat.pdf"},
+        headers={"Content-Disposition": "attachment;filename=chat.pdf"},
     )
 
 
@@ -125,7 +117,7 @@ async def download_db(user=Depends(get_admin_user)):
             status_code=status.HTTP_401_UNAUTHORIZED,
             detail=ERROR_MESSAGES.ACCESS_PROHIBITED,
         )
-    from apps.webui.internal.db import engine
+    from open_webui.apps.webui.internal.db import engine
 
     if engine.name != "sqlite":
         raise HTTPException(

backend/open_webui/apps/webui/utils.py

@@ -0,0 +1,170 @@
+import os
+import re
+import subprocess
+import sys
+from importlib import util
+import types
+import tempfile
+
+from open_webui.apps.webui.models.functions import Functions
+from open_webui.apps.webui.models.tools import Tools
+from open_webui.config import FUNCTIONS_DIR, TOOLS_DIR
+
+
+def extract_frontmatter(content):
+    """
+    Extract frontmatter as a dictionary from the provided content string.
+    """
+    frontmatter = {}
+    frontmatter_started = False
+    frontmatter_ended = False
+    frontmatter_pattern = re.compile(r"^\s*([a-z_]+):\s*(.*)\s*$", re.IGNORECASE)
+
+    try:
+        lines = content.splitlines()
+        if len(lines) < 1 or lines[0].strip() != '"""':
+            # The content doesn't start with triple quotes
+            return {}
+
+        frontmatter_started = True
+
+        for line in lines[1:]:
+            if '"""' in line:
+                if frontmatter_started:
+                    frontmatter_ended = True
+                    break
+
+            if frontmatter_started and not frontmatter_ended:
+                match = frontmatter_pattern.match(line)
+                if match:
+                    key, value = match.groups()
+                    frontmatter[key.strip()] = value.strip()
+
+    except Exception as e:
+        print(f"An error occurred: {e}")
+        return {}
+
+    return frontmatter
+
+
+def replace_imports(content):
+    """
+    Replace the import paths in the content.
+    """
+    replacements = {
+        "from utils": "from open_webui.utils",
+        "from apps": "from open_webui.apps",
+        "from main": "from open_webui.main",
+        "from config": "from open_webui.config",
+    }
+
+    for old, new in replacements.items():
+        content = content.replace(old, new)
+
+    return content
+
+
+def load_toolkit_module_by_id(toolkit_id, content=None):
+
+    if content is None:
+        tool = Tools.get_tool_by_id(toolkit_id)
+        if not tool:
+            raise Exception(f"Toolkit not found: {toolkit_id}")
+
+        content = tool.content
+
+        content = replace_imports(content)
+        Tools.update_tool_by_id(toolkit_id, {"content": content})
+    else:
+        frontmatter = extract_frontmatter(content)
+        # Install required packages found within the frontmatter
+        install_frontmatter_requirements(frontmatter.get("requirements", ""))
+
+    module_name = f"tool_{toolkit_id}"
+    module = types.ModuleType(module_name)
+    sys.modules[module_name] = module
+
+    # Create a temporary file and use it to define `__file__` so
+    # that it works as expected from the module's perspective.
+    temp_file = tempfile.NamedTemporaryFile(delete=False)
+
+    try:
+        with open(temp_file.name, "w", encoding="utf-8") as f:
+            f.write(content)
+        module.__dict__["__file__"] = temp_file.name
+
+        # Executing the modified content in the created module's namespace
+        exec(content, module.__dict__)
+        frontmatter = extract_frontmatter(content)
+        print(f"Loaded module: {module.__name__}")
+
+        # Create and return the object if the class 'Tools' is found in the module
+        if hasattr(module, "Tools"):
+            return module.Tools(), frontmatter
+        else:
+            raise Exception("No Tools class found in the module")
+    except Exception as e:
+        print(f"Error loading module: {toolkit_id}: {e}")
+        del sys.modules[module_name]  # Clean up
+        raise e
+    finally:
+        os.unlink(temp_file.name)
+
+
+def load_function_module_by_id(function_id, content=None):
+    if content is None:
+        function = Functions.get_function_by_id(function_id)
+        if not function:
+            raise Exception(f"Function not found: {function_id}")
+        content = function.content
+
+        content = replace_imports(content)
+        Functions.update_function_by_id(function_id, {"content": content})
+    else:
+        frontmatter = extract_frontmatter(content)
+        install_frontmatter_requirements(frontmatter.get("requirements", ""))
+
+    module_name = f"function_{function_id}"
+    module = types.ModuleType(module_name)
+    sys.modules[module_name] = module
+
+    # Create a temporary file and use it to define `__file__` so
+    # that it works as expected from the module's perspective.
+    temp_file = tempfile.NamedTemporaryFile(delete=False)
+    try:
+        with open(temp_file.name, "w", encoding="utf-8") as f:
+            f.write(content)
+        module.__dict__["__file__"] = temp_file.name
+
+        # Execute the modified content in the created module's namespace
+        exec(content, module.__dict__)
+        frontmatter = extract_frontmatter(content)
+        print(f"Loaded module: {module.__name__}")
+
+        # Create appropriate object based on available class type in the module
+        if hasattr(module, "Pipe"):
+            return module.Pipe(), "pipe", frontmatter
+        elif hasattr(module, "Filter"):
+            return module.Filter(), "filter", frontmatter
+        elif hasattr(module, "Action"):
+            return module.Action(), "action", frontmatter
+        else:
+            raise Exception("No Function class found in the module")
+    except Exception as e:
+        print(f"Error loading module: {function_id}: {e}")
+        del sys.modules[module_name]  # Cleanup by removing the module in case of error
+
+        Functions.update_function_by_id(function_id, {"is_active": False})
+        raise e
+    finally:
+        os.unlink(temp_file.name)
+
+
+def install_frontmatter_requirements(requirements):
+    if requirements:
+        req_list = [req.strip() for req in requirements.split(",")]
+        for req in req_list:
+            print(f"Installing requirement: {req}")
+            subprocess.check_call([sys.executable, "-m", "pip", "install", req])
+    else:
+        print("No requirements found in frontmatter.")

backend/open_webui/constants.py

@@ -100,3 +100,4 @@ class TASKS(str, Enum):
     EMOJI_GENERATION = "emoji_generation"
     QUERY_GENERATION = "query_generation"
     FUNCTION_CALLING = "function_calling"
+    MOA_RESPONSE_GENERATION = "moa_response_generation"

backend/open_webui/data/readme.txt

@@ -0,0 +1 @@
+pip install dir for backend files (db, documents, etc.)

backend/open_webui/env.py

@@ -0,0 +1,308 @@
+import importlib.metadata
+import json
+import logging
+import os
+import pkgutil
+import sys
+import shutil
+from pathlib import Path
+
+import markdown
+from bs4 import BeautifulSoup
+from open_webui.constants import ERROR_MESSAGES
+
+####################################
+# Load .env file
+####################################
+
+OPEN_WEBUI_DIR = Path(__file__).parent  # the path containing this file
+print(OPEN_WEBUI_DIR)
+
+BACKEND_DIR = OPEN_WEBUI_DIR.parent  # the path containing this file
+BASE_DIR = BACKEND_DIR.parent  # the path containing the backend/
+
+print(BACKEND_DIR)
+print(BASE_DIR)
+
+try:
+    from dotenv import find_dotenv, load_dotenv
+
+    load_dotenv(find_dotenv(str(BASE_DIR / ".env")))
+except ImportError:
+    print("dotenv not installed, skipping...")
+
+DOCKER = os.environ.get("DOCKER", "False").lower() == "true"
+
+# device type embedding models - "cpu" (default), "cuda" (nvidia gpu required) or "mps" (apple silicon) - choosing this right can lead to better performance
+USE_CUDA = os.environ.get("USE_CUDA_DOCKER", "false")
+
+if USE_CUDA.lower() == "true":
+    try:
+        import torch
+
+        assert torch.cuda.is_available(), "CUDA not available"
+        DEVICE_TYPE = "cuda"
+    except Exception as e:
+        cuda_error = (
+            "Error when testing CUDA but USE_CUDA_DOCKER is true. "
+            f"Resetting USE_CUDA_DOCKER to false: {e}"
+        )
+        os.environ["USE_CUDA_DOCKER"] = "false"
+        USE_CUDA = "false"
+        DEVICE_TYPE = "cpu"
+else:
+    DEVICE_TYPE = "cpu"
+
+
+####################################
+# LOGGING
+####################################
+
+log_levels = ["CRITICAL", "ERROR", "WARNING", "INFO", "DEBUG"]
+
+GLOBAL_LOG_LEVEL = os.environ.get("GLOBAL_LOG_LEVEL", "").upper()
+if GLOBAL_LOG_LEVEL in log_levels:
+    logging.basicConfig(stream=sys.stdout, level=GLOBAL_LOG_LEVEL, force=True)
+else:
+    GLOBAL_LOG_LEVEL = "INFO"
+
+log = logging.getLogger(__name__)
+log.info(f"GLOBAL_LOG_LEVEL: {GLOBAL_LOG_LEVEL}")
+
+if "cuda_error" in locals():
+    log.exception(cuda_error)
+
+log_sources = [
+    "AUDIO",
+    "COMFYUI",
+    "CONFIG",
+    "DB",
+    "IMAGES",
+    "MAIN",
+    "MODELS",
+    "OLLAMA",
+    "OPENAI",
+    "RAG",
+    "WEBHOOK",
+]
+
+SRC_LOG_LEVELS = {}
+
+for source in log_sources:
+    log_env_var = source + "_LOG_LEVEL"
+    SRC_LOG_LEVELS[source] = os.environ.get(log_env_var, "").upper()
+    if SRC_LOG_LEVELS[source] not in log_levels:
+        SRC_LOG_LEVELS[source] = GLOBAL_LOG_LEVEL
+    log.info(f"{log_env_var}: {SRC_LOG_LEVELS[source]}")
+
+log.setLevel(SRC_LOG_LEVELS["CONFIG"])
+
+
+WEBUI_NAME = os.environ.get("WEBUI_NAME", "Open WebUI")
+if WEBUI_NAME != "Open WebUI":
+    WEBUI_NAME += " (Open WebUI)"
+
+WEBUI_URL = os.environ.get("WEBUI_URL", "http://localhost:3000")
+
+WEBUI_FAVICON_URL = "https://openwebui.com/favicon.png"
+
+
+####################################
+# ENV (dev,test,prod)
+####################################
+
+ENV = os.environ.get("ENV", "dev")
+
+FROM_INIT_PY = os.environ.get("FROM_INIT_PY", "False").lower() == "true"
+
+if FROM_INIT_PY:
+    PACKAGE_DATA = {"version": importlib.metadata.version("open-webui")}
+else:
+    try:
+        PACKAGE_DATA = json.loads((BASE_DIR / "package.json").read_text())
+    except Exception:
+        PACKAGE_DATA = {"version": "0.0.0"}
+
+
+VERSION = PACKAGE_DATA["version"]
+
+
+# Function to parse each section
+def parse_section(section):
+    items = []
+    for li in section.find_all("li"):
+        # Extract raw HTML string
+        raw_html = str(li)
+
+        # Extract text without HTML tags
+        text = li.get_text(separator=" ", strip=True)
+
+        # Split into title and content
+        parts = text.split(": ", 1)
+        title = parts[0].strip() if len(parts) > 1 else ""
+        content = parts[1].strip() if len(parts) > 1 else text
+
+        items.append({"title": title, "content": content, "raw": raw_html})
+    return items
+
+
+try:
+    changelog_path = BASE_DIR / "CHANGELOG.md"
+    with open(str(changelog_path.absolute()), "r", encoding="utf8") as file:
+        changelog_content = file.read()
+
+except Exception:
+    changelog_content = (pkgutil.get_data("open_webui", "CHANGELOG.md") or b"").decode()
+
+
+# Convert markdown content to HTML
+html_content = markdown.markdown(changelog_content)
+
+# Parse the HTML content
+soup = BeautifulSoup(html_content, "html.parser")
+
+# Initialize JSON structure
+changelog_json = {}
+
+# Iterate over each version
+for version in soup.find_all("h2"):
+    version_number = version.get_text().strip().split(" - ")[0][1:-1]  # Remove brackets
+    date = version.get_text().strip().split(" - ")[1]
+
+    version_data = {"date": date}
+
+    # Find the next sibling that is a h3 tag (section title)
+    current = version.find_next_sibling()
+
+    while current and current.name != "h2":
+        if current.name == "h3":
+            section_title = current.get_text().lower()  # e.g., "added", "fixed"
+            section_items = parse_section(current.find_next_sibling("ul"))
+            version_data[section_title] = section_items
+
+        # Move to the next element
+        current = current.find_next_sibling()
+
+    changelog_json[version_number] = version_data
+
+
+CHANGELOG = changelog_json
+
+####################################
+# SAFE_MODE
+####################################
+
+SAFE_MODE = os.environ.get("SAFE_MODE", "false").lower() == "true"
+
+####################################
+# WEBUI_BUILD_HASH
+####################################
+
+WEBUI_BUILD_HASH = os.environ.get("WEBUI_BUILD_HASH", "dev-build")
+
+####################################
+# DATA/FRONTEND BUILD DIR
+####################################
+
+DATA_DIR = Path(os.getenv("DATA_DIR", BACKEND_DIR / "data")).resolve()
+
+if FROM_INIT_PY:
+    NEW_DATA_DIR = Path(os.getenv("DATA_DIR", OPEN_WEBUI_DIR / "data")).resolve()
+    NEW_DATA_DIR.mkdir(parents=True, exist_ok=True)
+
+    # Check if the data directory exists in the package directory
+    if DATA_DIR.exists() and DATA_DIR != NEW_DATA_DIR:
+        log.info(f"Moving {DATA_DIR} to {NEW_DATA_DIR}")
+        for item in DATA_DIR.iterdir():
+            dest = NEW_DATA_DIR / item.name
+            if item.is_dir():
+                shutil.copytree(item, dest, dirs_exist_ok=True)
+            else:
+                shutil.copy2(item, dest)
+
+    DATA_DIR = Path(os.getenv("DATA_DIR", OPEN_WEBUI_DIR / "data"))
+
+
+FONTS_DIR = Path(os.getenv("FONTS_DIR", OPEN_WEBUI_DIR / "static" / "fonts"))
+
+FRONTEND_BUILD_DIR = Path(os.getenv("FRONTEND_BUILD_DIR", BASE_DIR / "build")).resolve()
+
+if FROM_INIT_PY:
+    FRONTEND_BUILD_DIR = Path(
+        os.getenv("FRONTEND_BUILD_DIR", OPEN_WEBUI_DIR / "frontend")
+    ).resolve()
+
+
+RESET_CONFIG_ON_START = (
+    os.environ.get("RESET_CONFIG_ON_START", "False").lower() == "true"
+)
+
+if RESET_CONFIG_ON_START:
+    try:
+        os.remove(f"{DATA_DIR}/config.json")
+        with open(f"{DATA_DIR}/config.json", "w") as f:
+            f.write("{}")
+    except Exception:
+        pass
+
+####################################
+# Database
+####################################
+
+# Check if the file exists
+if os.path.exists(f"{DATA_DIR}/ollama.db"):
+    # Rename the file
+    os.rename(f"{DATA_DIR}/ollama.db", f"{DATA_DIR}/webui.db")
+    log.info("Database migrated from Ollama-WebUI successfully.")
+else:
+    pass
+
+DATABASE_URL = os.environ.get("DATABASE_URL", f"sqlite:///{DATA_DIR}/webui.db")
+
+# Replace the postgres:// with postgresql://
+if "postgres://" in DATABASE_URL:
+    DATABASE_URL = DATABASE_URL.replace("postgres://", "postgresql://")
+
+
+####################################
+# WEBUI_AUTH (Required for security)
+####################################
+
+WEBUI_AUTH = os.environ.get("WEBUI_AUTH", "True").lower() == "true"
+WEBUI_AUTH_TRUSTED_EMAIL_HEADER = os.environ.get(
+    "WEBUI_AUTH_TRUSTED_EMAIL_HEADER", None
+)
+WEBUI_AUTH_TRUSTED_NAME_HEADER = os.environ.get("WEBUI_AUTH_TRUSTED_NAME_HEADER", None)
+
+
+####################################
+# WEBUI_SECRET_KEY
+####################################
+
+WEBUI_SECRET_KEY = os.environ.get(
+    "WEBUI_SECRET_KEY",
+    os.environ.get(
+        "WEBUI_JWT_SECRET_KEY", "t0p-s3cr3t"
+    ),  # DEPRECATED: remove at next major version
+)
+
+WEBUI_SESSION_COOKIE_SAME_SITE = os.environ.get(
+    "WEBUI_SESSION_COOKIE_SAME_SITE",
+    os.environ.get("WEBUI_SESSION_COOKIE_SAME_SITE", "lax"),
+)
+
+WEBUI_SESSION_COOKIE_SECURE = os.environ.get(
+    "WEBUI_SESSION_COOKIE_SECURE",
+    os.environ.get("WEBUI_SESSION_COOKIE_SECURE", "false").lower() == "true",
+)
+
+if WEBUI_AUTH and WEBUI_SECRET_KEY == "":
+    raise ValueError(ERROR_MESSAGES.ENV_VAR_NOT_FOUND)
+
+ENABLE_WEBSOCKET_SUPPORT = (
+    os.environ.get("ENABLE_WEBSOCKET_SUPPORT", "True").lower() == "true"
+)
+
+WEBSOCKET_MANAGER = os.environ.get("WEBSOCKET_MANAGER", "")
+
+WEBSOCKET_REDIS_URL = os.environ.get("WEBSOCKET_REDIS_URL", "redis://localhost:6379/0")

backend/open_webui/migrations/env.py

@@ -1,24 +1,9 @@
-import os
 from logging.config import fileConfig
 
-from sqlalchemy import engine_from_config
-from sqlalchemy import pool
-
 from alembic import context
-
-from apps.webui.models.auths import Auth
-from apps.webui.models.chats import Chat
-from apps.webui.models.documents import Document
-from apps.webui.models.memories import Memory
-from apps.webui.models.models import Model
-from apps.webui.models.prompts import Prompt
-from apps.webui.models.tags import Tag, ChatIdTag
-from apps.webui.models.tools import Tool
-from apps.webui.models.users import User
-from apps.webui.models.files import File
-from apps.webui.models.functions import Function
-
-from config import DATABASE_URL
+from open_webui.apps.webui.models.auths import Auth
+from open_webui.env import DATABASE_URL
+from sqlalchemy import engine_from_config, pool
 
 # this is the Alembic Config object, which provides
 # access to the values within the .ini file in use.

backend/open_webui/migrations/script.py.mako

@@ -9,7 +9,7 @@ from typing import Sequence, Union
 
 from alembic import op
 import sqlalchemy as sa
-import apps.webui.internal.db
+import open_webui.apps.webui.internal.db
 ${imports if imports else ""}
 
 # revision identifiers, used by Alembic.