Merge pull request #9190 from open-webui/dev

0.5.8
refac: styling
2025-02-05 01:16:35 -08:00 · 2025-02-05 01:14:46 -08:00 · 2025-02-05 01:09:23 -08:00 · 2025-02-05 01:08:39 -08:00 · 2025-02-05 01:07:38 -08:00 · 2025-02-05 10:05:38 +01:00
4107 changed files with 25288 additions and 5952 deletions
@@ -0,0 +1,25 @@
+# Codespell configuration is within pyproject.toml
+---
+name: Codespell
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+    branches: [main]
+
+permissions:
+  contents: read
+
+jobs:
+  codespell:
+    name: Check for spelling errors
+    runs-on: ubuntu-latest
+
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+      - name: Annotate locations with typos
+        uses: codespell-project/codespell-problem-matcher@v1
+      - name: Codespell
+        uses: codespell-project/actions-codespell@v2
@@ -52,6 +52,8 @@ jobs:

      - name: Cypress run
        uses: cypress-io/github-action@v6
+        env:
+          LIBGL_ALWAYS_SOFTWARE: 1
        with:
          browser: chrome
          wait-on: 'http://localhost:3000'
@@ -5,6 +5,133 @@ 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).

+## [0.5.8] - 2025-02-05
+
+### Added
+
+- **🖥️ Code Interpreter**: Models can now execute code in real time to refine their answers dynamically, running securely within a sandboxed browser environment using Pyodide. Perfect for calculations, data analysis, and AI-assisted coding tasks!
+- **💬 Redesigned Chat Input UI**: Enjoy a sleeker and more intuitive message input with improved feature selection, making it easier than ever to toggle tools, enable search, and interact with AI seamlessly.
+- **🛠️ Native Tool Calling Support (Experimental)**: Supported models can now call tools natively, reducing query latency and improving contextual responses. More enhancements coming soon!
+- **🔗 Exa Search Engine Integration**: A new search provider has been added, allowing users to retrieve up-to-date and relevant information without leaving the chat interface.
+- **🌍 Localized Dates & Times**: Date and time formats now match your system locale, ensuring a more natural, region-specific experience.
+- **📎 User Headers for External Embedding APIs**: API calls to external embedding services now include user-related headers.
+- **🌍 "Always On" Web Search Toggle**: A new option under Settings > Interface allows users to enable Web Search by default—transform Open WebUI into your go-to search engine, ensuring AI-powered results with every query.
+- **🚀 General Performance & Stability**: Significant improvements across the platform for a faster, more reliable experience.
+- **🖼️ UI/UX Enhancements**: Numerous design refinements improving readability, responsiveness, and accessibility.
+- **🌍 Improved Translations**: Chinese, Korean, French, Ukrainian and Serbian translations have been updated with refined terminologies for better clarity.
+
+### Fixed
+
+- **🔄 OAuth Name Field Fallback**: Resolves OAuth login failures by using the email field as a fallback when a name is missing.
+- **🔑 Google Drive Credentials Restriction**: Ensures only authenticated users can access Google Drive credentials for enhanced security.
+- **🌐 DuckDuckGo Search Rate Limit Handling**: Fixes issues where users would encounter 202 errors due to rate limits when using DuckDuckGo for web search.
+- **📁 File Upload Permission Indicator**: Users are now notified when they lack permission to upload files, improving clarity on system restrictions.
+- **🔧 Max Tokens Issue**: Fixes cases where 'max_tokens' were not applied correctly, ensuring proper model behavior.
+- **🔍 Validation for RAG Web Search URLs**: Filters out invalid or unsupported URLs when using web-based retrieval augmentation.
+- **🖋️ Title Generation Bug**: Fixes inconsistencies in title generation, ensuring proper chat organization.
+
+### Removed
+
+- **⚡ Deprecated Non-Web Worker Pyodide Execution**: Moves entirely to browser sandboxing for better performance and security.
+
+## [0.5.7] - 2025-01-23
+
+### Added
+
+- **🌍 Enhanced Internationalization (i18n)**: Refined and expanded translations for greater global accessibility and a smoother experience for international users.
+
+### Fixed
+
+- **🔗 Connection Model ID Resolution**: Resolved an issue preventing model IDs from registering in connections.
+- **💡 Prefix ID for Ollama Connections**: Fixed a bug where prefix IDs in Ollama connections were non-functional.
+- **🔧 Ollama Model Enable/Disable Functionality**: Addressed the issue of enable/disable toggles not working for Ollama base models.
+- **🔒 RBAC Permissions for Tools and Models**: Corrected incorrect Role-Based Access Control (RBAC) permissions for tools and models, ensuring that users now only access features according to their assigned privileges, enhancing security and role clarity.
+
+## [0.5.6] - 2025-01-22
+
+### Added
+
+- **🧠 Effortful Reasoning Control for OpenAI Models**: Introduced the reasoning_effort parameter in chat controls for supported OpenAI models, enabling users to fine-tune how much cognitive effort a model dedicates to its responses, offering greater customization for complex queries and reasoning tasks.
+
+### Fixed
+
+- **🔄 Chat Controls Loading UI Bug**: Resolved an issue where collapsible chat controls appeared as "loading," ensuring a smoother and more intuitive user experience for managing chat settings.
+
+### Changed
+
+- **🔧 Updated Ollama Model Creation**: Revamped the Ollama model creation method to align with their new JSON payload format, ensuring seamless compatibility and more efficient model setup workflows.
+
+## [0.5.5] - 2025-01-22
+
+### Added
+
+- **🤔 Native 'Think' Tag Support**: Introduced the new 'think' tag support that visually displays how long the model is thinking, omitting the reasoning content itself until the next turn. Ideal for creating a more streamlined and focused interaction experience.
+- **🖼️ Toggle Image Generation On/Off**: In the chat input menu, you can now easily toggle image generation before initiating chats, providing greater control and flexibility to suit your needs.
+- **🔒 Chat Controls Permissions**: Admins can now disable chat controls access for users, offering tighter management and customization over user interactions.
+- **🔍 Web Search & Image Generation Permissions**: Easily disable web search and image generation for specific users, improving workflow governance and security for certain environments.
+- **🗂️ S3 and GCS Storage Provider Support**: Scaled deployments now benefit from expanded storage options with Amazon S3 and Google Cloud Storage seamlessly integrated as providers.
+- **🎨 Enhanced Model Management**: Reintroduced the ability to download and delete models directly in the admin models settings page to minimize user confusion and aid efficient model management.
+- **🔗 Improved Connection Handling**: Enhanced backend to smoothly handle multiple identical base URLs, allowing more flexible multi-instance configurations with fewer hiccups.
+- **✨ General UI/UX Refinements**: Numerous tweaks across the WebUI make navigation and usability even more user-friendly and intuitive.
+- **🌍 Translation Enhancements**: Various translation updates ensure smoother and more polished interactions for international users.
+
+### Fixed
+
+- **⚡ MPS Functionality for Mac Users**: Fixed MPS support, ensuring smooth performance and compatibility for Mac users leveraging MPS.
+- **📡 Ollama Connection Management**: Resolved the issue where deleting all Ollama connections prevented adding new ones.
+
+### Changed
+
+- **⚙️ General Stability Refac**: Backend refactoring delivers a more stable, robust platform.
+- **🖥️ Desktop App Preparations**: Ongoing work to support the upcoming Open WebUI desktop app. Follow our progress and updates here: https://github.com/open-webui/desktop
+
+## [0.5.4] - 2025-01-05
+
+### Added
+
+- **🔄 Clone Shared Chats**: Effortlessly clone shared chats to save time and streamline collaboration, perfect for reusing insightful discussions or custom setups.
+- **📣 Native Notifications for Channel Messages**: Stay informed with integrated desktop notifications for channel messages, ensuring you never miss important updates while multitasking.
+- **🔥 Torch MPS Support**: MPS support for Mac users when Open WebUI is installed directly, offering better performance and compatibility for AI workloads.
+- **🌍 Enhanced Translations**: Small improvements to various translations, ensuring a smoother global user experience.
+
+### Fixed
+
+- **🖼️ Image-Only Messages in Channels**: You can now send images without accompanying text or content in channels.
+- **❌ Proper Exception Handling**: Enhanced error feedback by ensuring exceptions are raised clearly, reducing confusion and promoting smoother debugging.
+- **🔍 RAG Query Generation Restored**: Fixed query generation issues for Retrieval-Augmented Generation, improving retrieval accuracy and ensuring seamless functionality.
+- **📩 MOA Response Functionality Fixed**: Addressed an error with the MOA response generation feature.
+- **💬 Channel Thread Loading with 50+ Messages**: Resolved an issue where channel threads stalled when exceeding 50 messages, ensuring smooth navigation in active discussions.
+- **🔑 API Endpoint Restrictions Resolution**: Fixed a critical bug where the 'API_KEY_ALLOWED_ENDPOINTS' setting was not functioning as intended, ensuring API access is limited to specified endpoints for enhanced security.
+- **🛠️ Action Functions Restored**: Corrected an issue preventing action functions from working, restoring their utility for customized automations and workflows.
+- **📂 Temporary Chat JSON Export Fix**: Resolved a bug blocking temporary chats from being exported in JSON format, ensuring seamless data portability.
+
+### Changed
+
+- **🎛️ Sidebar UI Tweaks**: Chat folders, including pinned folders, now display below the Chats section for better organization; the "New Folder" button has been relocated to the Chats section for a more intuitive workflow.
+- **🏗️ Real-Time Save Disabled by Default**: The 'ENABLE_REALTIME_CHAT_SAVE' setting is now off by default, boosting response speed for users who prioritize performance in high-paced workflows or less critical scenarios.
+- **🎤 Audio Input Echo Cancellation**: Audio input now features echo cancellation enabled by default, reducing audio feedback for improved clarity during conversations or voice-based interactions.
+- **🔧 General Reliability Improvements**: Numerous under-the-hood enhancements have been made to improve platform stability, boost overall performance, and ensure a more seamless, dependable experience across workflows.
+
+## [0.5.3] - 2024-12-31
+
+### Added
+
+- **💬 Channel Reactions with Built-In Emoji Picker**: Easily express yourself in channel threads and messages with reactions, featuring an intuitive built-in emoji picker for seamless selection.
+- **🧵 Threads for Channels**: Organize discussions within channels by creating threads, improving clarity and fostering focused conversations.
+- **🔄 Reset Button for SVG Pan/Zoom**: Added a handy reset button to SVG Pan/Zoom, allowing users to quickly return diagrams or visuals to their default state without hassle.
+- **⚡ Realtime Chat Save Environment Variable**: Introduced the ENABLE_REALTIME_CHAT_SAVE environment variable. Choose between faster responses by disabling realtime chat saving or ensuring chunk-by-chunk data persistency for critical operations.
+- **🌍 Translation Enhancements**: Updated and refined translations across multiple languages, providing a smoother experience for international users.
+- **📚 Improved Documentation**: Expanded documentation on functions, including clearer guidance on function plugins and detailed instructions for migrating to v0.5. This ensures users can adapt and harness new updates more effectively. (https://docs.openwebui.com/features/plugin/)
+
+### Fixed
+
+- **🛠️ Ollama Parameters Respected**: Resolved an issue where input parameters for Ollama were being ignored, ensuring precise and consistent model behavior.
+- **🔧 Function Plugin Outlet Hook Reliability**: Fixed a bug causing issues with 'event_emitter' and outlet hooks in filter function plugins, guaranteeing smoother operation within custom extensions.
+- **🖋️ Weird Custom Status Descriptions**: Adjusted the formatting and functionality for custom user statuses, ensuring they display correctly and intuitively.
+- **🔗 Restored API Functionality**: Fixed a critical issue where APIs were not operational for certain configurations, ensuring uninterrupted access.
+- **⏳ Custom Pipe Function Completion**: Resolved an issue where chats using specific custom pipe function plugins weren’t finishing properly, restoring consistent chat workflows.
+- **✅ General Stability Enhancements**: Implemented various under-the-hood improvements to boost overall reliability, ensuring smoother and more consistent performance across the WebUI.
+
 ## [0.5.2] - 2024-12-26

 ### Added
@@ -2,7 +2,7 @@

 ## Our Pledge

-As members, contributors, and leaders of this community, we pledge to make participation in our open-source project a harassment-free experience for everyone, regardless of age, body size, visible or invisible disability, ethnicity, sex characteristics, gender identity and expression, level of experience, education, socio-economic status, nationality, personal appearance, race, religion, or sexual identity and orientation.
+As members, contributors, and leaders of this community, we pledge to make participation in our open-source project a harassment-free experience for everyone, regardless of age, body size, visible or invisible disability, ethnicity, sex characteristics, gender identity and expression, level of experience, education, socioeconomic status, nationality, personal appearance, race, religion, or sexual identity and orientation.

 We are committed to creating and maintaining an open, respectful, and professional environment where positive contributions and meaningful discussions can flourish. By participating in this project, you agree to uphold these values and align your behavior to the standards outlined in this Code of Conduct.

@@ -1,21 +1,27 @@
-MIT License
+Copyright (c) 2023-2025 Timothy Jaeryang Baek
+All rights reserved.

-Copyright (c) 2023 Timothy Jaeryang Baek
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are met:

-Permission is hereby granted, free of charge, to any person obtaining a copy
-of this software and associated documentation files (the "Software"), to deal
-in the Software without restriction, including without limitation the rights
-to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
-copies of the Software, and to permit persons to whom the Software is
-furnished to do so, subject to the following conditions:
+1. Redistributions of source code must retain the above copyright notice, this
+   list of conditions and the following disclaimer.

-The above copyright notice and this permission notice shall be included in all
-copies or substantial portions of the Software.
+2. Redistributions in binary form must reproduce the above copyright notice,
+   this list of conditions and the following disclaimer in the documentation
+   and/or other materials provided with the distribution.

-THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
-AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
-LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
-OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
-SOFTWARE.
+3. Neither the name of the copyright holder nor the names of its
+   contributors may be used to endorse or promote products derived from
+   this software without specific prior written permission.
+
+THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
+FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
+CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
+OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
@@ -11,7 +11,9 @@
 [![Discord](https://img.shields.io/badge/Discord-Open_WebUI-blue?logo=discord&logoColor=white)](https://discord.gg/5rJgQTnV4s)
 [![](https://img.shields.io/static/v1?label=Sponsor&message=%E2%9D%A4&logo=GitHub&color=%23fe8e86)](https://github.com/sponsors/tjbck)

-Open WebUI is an [extensible](https://github.com/open-webui/pipelines), feature-rich, and user-friendly self-hosted WebUI designed to operate entirely offline. It supports various LLM runners, including Ollama and OpenAI-compatible APIs. For more information, be sure to check out our [Open WebUI Documentation](https://docs.openwebui.com/).
+**Open WebUI is an [extensible](https://docs.openwebui.com/features/plugin/), feature-rich, and user-friendly self-hosted AI platform designed to operate entirely offline.** It supports various LLM runners like **Ollama** and **OpenAI-compatible APIs**, with **built-in inference engine** for RAG, making it a **powerful AI deployment solution**.
+
+For more information, be sure to check out our [Open WebUI Documentation](https://docs.openwebui.com/).

 ![Open WebUI Demo](./demo.gif)

@@ -185,13 +187,21 @@ If you want to try out the latest bleeding-edge features and are okay with occas
 docker run -d -p 3000:8080 -v open-webui:/app/backend/data --name open-webui --add-host=host.docker.internal:host-gateway --restart always ghcr.io/open-webui/open-webui:dev
 ```

+### Offline Mode
+
+If you are running Open WebUI in an offline environment, you can set the `HF_HUB_OFFLINE` environment variable to `1` to prevent attempts to download models from the internet.
+
+```bash
+export HF_HUB_OFFLINE=1
+```
+
 ## What's Next? 🌟

 Discover upcoming features on our roadmap in the [Open WebUI Documentation](https://docs.openwebui.com/roadmap/).

 ## License 📜

-This project is licensed under the [MIT License](LICENSE) - see the [LICENSE](LICENSE) file for details. 📄
+This project is licensed under the [BSD-3-Clause License](LICENSE) - see the [LICENSE](LICENSE) file for details. 📄

 ## Support 💬

@@ -5,12 +5,31 @@ from pathlib import Path

 import typer
 import uvicorn
+from typing import Optional
+from typing_extensions import Annotated

 app = typer.Typer()

 KEY_FILE = Path.cwd() / ".webui_secret_key"


+def version_callback(value: bool):
+    if value:
+        from open_webui.env import VERSION
+
+        typer.echo(f"Open WebUI version: {VERSION}")
+        raise typer.Exit()
+
+
+@app.command()
+def main(
+    version: Annotated[
+        Optional[bool], typer.Option("--version", callback=version_callback)
+    ] = None,
+):
+    pass
+
+
@app.command()
 def serve(
    host: str = "0.0.0.0",
@@ -9,22 +9,22 @@ from urllib.parse import urlparse

 import chromadb
 import requests
-import yaml
-from open_webui.internal.db import Base, get_db
+from pydantic import BaseModel
+from sqlalchemy import JSON, Column, DateTime, Integer, func
+
 from open_webui.env import (
-    OPEN_WEBUI_DIR,
    DATA_DIR,
+    DATABASE_URL,
    ENV,
    FRONTEND_BUILD_DIR,
+    OFFLINE_MODE,
+    OPEN_WEBUI_DIR,
    WEBUI_AUTH,
    WEBUI_FAVICON_URL,
    WEBUI_NAME,
    log,
-    DATABASE_URL,
-    OFFLINE_MODE,
 )
-from pydantic import BaseModel
-from sqlalchemy import JSON, Column, DateTime, Integer, func
+from open_webui.internal.db import Base, get_db


 class EndpointFilter(logging.Filter):
@@ -362,6 +362,30 @@ MICROSOFT_REDIRECT_URI = PersistentConfig(
    os.environ.get("MICROSOFT_REDIRECT_URI", ""),
 )

+GITHUB_CLIENT_ID = PersistentConfig(
+    "GITHUB_CLIENT_ID",
+    "oauth.github.client_id",
+    os.environ.get("GITHUB_CLIENT_ID", ""),
+)
+
+GITHUB_CLIENT_SECRET = PersistentConfig(
+    "GITHUB_CLIENT_SECRET",
+    "oauth.github.client_secret",
+    os.environ.get("GITHUB_CLIENT_SECRET", ""),
+)
+
+GITHUB_CLIENT_SCOPE = PersistentConfig(
+    "GITHUB_CLIENT_SCOPE",
+    "oauth.github.scope",
+    os.environ.get("GITHUB_CLIENT_SCOPE", "user:email"),
+)
+
+GITHUB_CLIENT_REDIRECT_URI = PersistentConfig(
+    "GITHUB_CLIENT_REDIRECT_URI",
+    "oauth.github.redirect_uri",
+    os.environ.get("GITHUB_CLIENT_REDIRECT_URI", ""),
+)
+
 OAUTH_CLIENT_ID = PersistentConfig(
    "OAUTH_CLIENT_ID",
    "oauth.oidc.client_id",
@@ -468,12 +492,20 @@ OAUTH_ALLOWED_DOMAINS = PersistentConfig(
 def load_oauth_providers():
    OAUTH_PROVIDERS.clear()
    if GOOGLE_CLIENT_ID.value and GOOGLE_CLIENT_SECRET.value:
+
+        def google_oauth_register(client):
+            client.register(
+                name="google",
+                client_id=GOOGLE_CLIENT_ID.value,
+                client_secret=GOOGLE_CLIENT_SECRET.value,
+                server_metadata_url="https://accounts.google.com/.well-known/openid-configuration",
+                client_kwargs={"scope": GOOGLE_OAUTH_SCOPE.value},
+                redirect_uri=GOOGLE_REDIRECT_URI.value,
+            )
+
        OAUTH_PROVIDERS["google"] = {
-            "client_id": GOOGLE_CLIENT_ID.value,
-            "client_secret": GOOGLE_CLIENT_SECRET.value,
-            "server_metadata_url": "https://accounts.google.com/.well-known/openid-configuration",
-            "scope": GOOGLE_OAUTH_SCOPE.value,
            "redirect_uri": GOOGLE_REDIRECT_URI.value,
+            "register": google_oauth_register,
        }

    if (
@@ -481,12 +513,44 @@ def load_oauth_providers():
        and MICROSOFT_CLIENT_SECRET.value
        and MICROSOFT_CLIENT_TENANT_ID.value
    ):
+
+        def microsoft_oauth_register(client):
+            client.register(
+                name="microsoft",
+                client_id=MICROSOFT_CLIENT_ID.value,
+                client_secret=MICROSOFT_CLIENT_SECRET.value,
+                server_metadata_url=f"https://login.microsoftonline.com/{MICROSOFT_CLIENT_TENANT_ID.value}/v2.0/.well-known/openid-configuration",
+                client_kwargs={
+                    "scope": MICROSOFT_OAUTH_SCOPE.value,
+                },
+                redirect_uri=MICROSOFT_REDIRECT_URI.value,
+            )
+
        OAUTH_PROVIDERS["microsoft"] = {
-            "client_id": MICROSOFT_CLIENT_ID.value,
-            "client_secret": MICROSOFT_CLIENT_SECRET.value,
-            "server_metadata_url": f"https://login.microsoftonline.com/{MICROSOFT_CLIENT_TENANT_ID.value}/v2.0/.well-known/openid-configuration",
-            "scope": MICROSOFT_OAUTH_SCOPE.value,
            "redirect_uri": MICROSOFT_REDIRECT_URI.value,
+            "picture_url": "https://graph.microsoft.com/v1.0/me/photo/$value",
+            "register": microsoft_oauth_register,
+        }
+
+    if GITHUB_CLIENT_ID.value and GITHUB_CLIENT_SECRET.value:
+
+        def github_oauth_register(client):
+            client.register(
+                name="github",
+                client_id=GITHUB_CLIENT_ID.value,
+                client_secret=GITHUB_CLIENT_SECRET.value,
+                access_token_url="https://github.com/login/oauth/access_token",
+                authorize_url="https://github.com/login/oauth/authorize",
+                api_base_url="https://api.github.com",
+                userinfo_endpoint="https://api.github.com/user",
+                client_kwargs={"scope": GITHUB_CLIENT_SCOPE.value},
+                redirect_uri=GITHUB_CLIENT_REDIRECT_URI.value,
+            )
+
+        OAUTH_PROVIDERS["github"] = {
+            "redirect_uri": GITHUB_CLIENT_REDIRECT_URI.value,
+            "register": github_oauth_register,
+            "sub_claim": "id",
        }

    if (
@@ -494,13 +558,23 @@ def load_oauth_providers():
        and OAUTH_CLIENT_SECRET.value
        and OPENID_PROVIDER_URL.value
    ):
+
+        def oidc_oauth_register(client):
+            client.register(
+                name="oidc",
+                client_id=OAUTH_CLIENT_ID.value,
+                client_secret=OAUTH_CLIENT_SECRET.value,
+                server_metadata_url=OPENID_PROVIDER_URL.value,
+                client_kwargs={
+                    "scope": OAUTH_SCOPES.value,
+                },
+                redirect_uri=OPENID_REDIRECT_URI.value,
+            )
+
        OAUTH_PROVIDERS["oidc"] = {
-            "client_id": OAUTH_CLIENT_ID.value,
-            "client_secret": OAUTH_CLIENT_SECRET.value,
-            "server_metadata_url": OPENID_PROVIDER_URL.value,
-            "scope": OAUTH_SCOPES.value,
            "name": OAUTH_PROVIDER_NAME.value,
            "redirect_uri": OPENID_REDIRECT_URI.value,
+            "register": oidc_oauth_register,
        }


@@ -580,7 +654,7 @@ if CUSTOM_NAME:
 # STORAGE PROVIDER
 ####################################

-STORAGE_PROVIDER = os.environ.get("STORAGE_PROVIDER", "")  # defaults to local, s3
+STORAGE_PROVIDER = os.environ.get("STORAGE_PROVIDER", "local")  # defaults to local, s3

 S3_ACCESS_KEY_ID = os.environ.get("S3_ACCESS_KEY_ID", None)
 S3_SECRET_ACCESS_KEY = os.environ.get("S3_SECRET_ACCESS_KEY", None)
@@ -588,6 +662,11 @@ S3_REGION_NAME = os.environ.get("S3_REGION_NAME", None)
 S3_BUCKET_NAME = os.environ.get("S3_BUCKET_NAME", None)
 S3_ENDPOINT_URL = os.environ.get("S3_ENDPOINT_URL", None)

+GCS_BUCKET_NAME = os.environ.get("GCS_BUCKET_NAME", None)
+GOOGLE_APPLICATION_CREDENTIALS_JSON = os.environ.get(
+    "GOOGLE_APPLICATION_CREDENTIALS_JSON", None
+)
+
 ####################################
 # File Upload DIR
 ####################################
@@ -819,6 +898,10 @@ USER_PERMISSIONS_WORKSPACE_TOOLS_ACCESS = (
    os.environ.get("USER_PERMISSIONS_WORKSPACE_TOOLS_ACCESS", "False").lower() == "true"
 )

+USER_PERMISSIONS_CHAT_CONTROLS = (
+    os.environ.get("USER_PERMISSIONS_CHAT_CONTROLS", "True").lower() == "true"
+)
+
 USER_PERMISSIONS_CHAT_FILE_UPLOAD = (
    os.environ.get("USER_PERMISSIONS_CHAT_FILE_UPLOAD", "True").lower() == "true"
 )
@@ -835,23 +918,46 @@ USER_PERMISSIONS_CHAT_TEMPORARY = (
    os.environ.get("USER_PERMISSIONS_CHAT_TEMPORARY", "True").lower() == "true"
 )

+USER_PERMISSIONS_FEATURES_WEB_SEARCH = (
+    os.environ.get("USER_PERMISSIONS_FEATURES_WEB_SEARCH", "True").lower() == "true"
+)
+
+USER_PERMISSIONS_FEATURES_IMAGE_GENERATION = (
+    os.environ.get("USER_PERMISSIONS_FEATURES_IMAGE_GENERATION", "True").lower()
+    == "true"
+)
+
+USER_PERMISSIONS_FEATURES_CODE_INTERPRETER = (
+    os.environ.get("USER_PERMISSIONS_FEATURES_CODE_INTERPRETER", "True").lower()
+    == "true"
+)
+
+
+DEFAULT_USER_PERMISSIONS = {
+    "workspace": {
+        "models": USER_PERMISSIONS_WORKSPACE_MODELS_ACCESS,
+        "knowledge": USER_PERMISSIONS_WORKSPACE_KNOWLEDGE_ACCESS,
+        "prompts": USER_PERMISSIONS_WORKSPACE_PROMPTS_ACCESS,
+        "tools": USER_PERMISSIONS_WORKSPACE_TOOLS_ACCESS,
+    },
+    "chat": {
+        "controls": USER_PERMISSIONS_CHAT_CONTROLS,
+        "file_upload": USER_PERMISSIONS_CHAT_FILE_UPLOAD,
+        "delete": USER_PERMISSIONS_CHAT_DELETE,
+        "edit": USER_PERMISSIONS_CHAT_EDIT,
+        "temporary": USER_PERMISSIONS_CHAT_TEMPORARY,
+    },
+    "features": {
+        "web_search": USER_PERMISSIONS_FEATURES_WEB_SEARCH,
+        "image_generation": USER_PERMISSIONS_FEATURES_IMAGE_GENERATION,
+        "code_interpreter": USER_PERMISSIONS_FEATURES_CODE_INTERPRETER,
+    },
+}
+
 USER_PERMISSIONS = PersistentConfig(
    "USER_PERMISSIONS",
    "user.permissions",
-    {
-        "workspace": {
-            "models": USER_PERMISSIONS_WORKSPACE_MODELS_ACCESS,
-            "knowledge": USER_PERMISSIONS_WORKSPACE_KNOWLEDGE_ACCESS,
-            "prompts": USER_PERMISSIONS_WORKSPACE_PROMPTS_ACCESS,
-            "tools": USER_PERMISSIONS_WORKSPACE_TOOLS_ACCESS,
-        },
-        "chat": {
-            "file_upload": USER_PERMISSIONS_CHAT_FILE_UPLOAD,
-            "delete": USER_PERMISSIONS_CHAT_DELETE,
-            "edit": USER_PERMISSIONS_CHAT_EDIT,
-            "temporary": USER_PERMISSIONS_CHAT_TEMPORARY,
-        },
-    },
+    DEFAULT_USER_PERMISSIONS,
 )

 ENABLE_CHANNELS = PersistentConfig(
@@ -995,21 +1101,27 @@ TITLE_GENERATION_PROMPT_TEMPLATE = PersistentConfig(
    os.environ.get("TITLE_GENERATION_PROMPT_TEMPLATE", ""),
 )

-DEFAULT_TITLE_GENERATION_PROMPT_TEMPLATE = """Create a concise, 3-5 word title with an emoji as a title for the chat history, in the given language. Suitable Emojis for the summary can be used to enhance understanding but avoid quotation marks or special formatting. RESPOND ONLY WITH THE TITLE TEXT.
-
-Examples of titles:
-📉 Stock Market Trends
-🍪 Perfect Chocolate Chip Recipe
-Evolution of Music Streaming
-Remote Work Productivity Tips
-Artificial Intelligence in Healthcare
-🎮 Video Game Development Insights
-
+DEFAULT_TITLE_GENERATION_PROMPT_TEMPLATE = """### Task:
+Generate a concise, 3-5 word title with an emoji summarizing the chat history.
+### Guidelines:
+- The title should clearly represent the main theme or subject of the conversation.
+- Use emojis that enhance understanding of the topic, but avoid quotation marks or special formatting.
+- Write the title in the chat's primary language; default to English if multilingual.
+- Prioritize accuracy over excessive creativity; keep it clear and simple.
+### Output:
+JSON format: { "title": "your concise title here" }
+### Examples:
+- { "title": "📉 Stock Market Trends" },
+- { "title": "🍪 Perfect Chocolate Chip Recipe" },
+- { "title": "Evolution of Music Streaming" },
+- { "title": "Remote Work Productivity Tips" },
+- { "title": "Artificial Intelligence in Healthcare" },
+- { "title": "🎮 Video Game Development Insights" }
+### Chat History:
 <chat_history>
 {{MESSAGES:END:2}}
 </chat_history>"""

-
 TAGS_GENERATION_PROMPT_TEMPLATE = PersistentConfig(
    "TAGS_GENERATION_PROMPT_TEMPLATE",
    "task.tags.prompt_template",
@@ -1034,6 +1146,32 @@ JSON format: { "tags": ["tag1", "tag2", "tag3"] }
 {{MESSAGES:END:6}}
 </chat_history>"""

+IMAGE_PROMPT_GENERATION_PROMPT_TEMPLATE = PersistentConfig(
+    "IMAGE_PROMPT_GENERATION_PROMPT_TEMPLATE",
+    "task.image.prompt_template",
+    os.environ.get("IMAGE_PROMPT_GENERATION_PROMPT_TEMPLATE", ""),
+)
+
+DEFAULT_IMAGE_PROMPT_GENERATION_PROMPT_TEMPLATE = """### Task:
+Generate a detailed prompt for am image generation task based on the given language and context. Describe the image as if you were explaining it to someone who cannot see it. Include relevant details, colors, shapes, and any other important elements.
+
+### Guidelines:
+- Be descriptive and detailed, focusing on the most important aspects of the image.
+- Avoid making assumptions or adding information not present in the image.
+- Use the chat's primary language; default to English if multilingual.
+- If the image is too complex, focus on the most prominent elements.
+
+### Output:
+Strictly return in JSON format:
+{
+    "prompt": "Your detailed description here."
+}
+
+### Chat History:
+<chat_history>
+{{MESSAGES:END:6}}
+</chat_history>"""
+
 ENABLE_TAGS_GENERATION = PersistentConfig(
    "ENABLE_TAGS_GENERATION",
    "task.tags.enable",
@@ -1152,7 +1290,28 @@ TOOLS_FUNCTION_CALLING_PROMPT_TEMPLATE = PersistentConfig(
 )


-DEFAULT_TOOLS_FUNCTION_CALLING_PROMPT_TEMPLATE = """Available Tools: {{TOOLS}}\nReturn an empty string if no tools match the query. If a function tool matches, construct and return a JSON object in the format {\"name\": \"functionName\", \"parameters\": {\"requiredFunctionParamKey\": \"requiredFunctionParamValue\"}} using the appropriate tool and its parameters. Only return the object and limit the response to the JSON object without additional text."""
+DEFAULT_TOOLS_FUNCTION_CALLING_PROMPT_TEMPLATE = """Available Tools: {{TOOLS}}
+
+Your task is to choose and return the correct tool(s) from the list of available tools based on the query. Follow these guidelines:
+
+- Return only the JSON object, without any additional text or explanation.
+
+- If no tools match the query, return an empty array: 
+   {
+     "tool_calls": []
+   }
+
+- If one or more tools match the query, construct a JSON response containing a "tool_calls" array with objects that include:
+   - "name": The tool's name.
+   - "parameters": A dictionary of required parameters and their corresponding values.
+
+The format for the JSON response is strictly:
+{
+  "tool_calls": [
+    {"name": "toolName1", "parameters": {"key1": "value1"}},
+    {"name": "toolName2", "parameters": {"key2": "value2"}}
+  ]
+}"""


 DEFAULT_EMOJI_GENERATION_PROMPT_TEMPLATE = """Your task is to reflect the speaker's likely facial expression through a fitting emoji. Interpret emotions from the message and reflect their facial expression using fitting, diverse emojis (e.g., 😊, 😢, 😡, 😱).
@@ -1165,6 +1324,24 @@ Your task is to synthesize these responses into a single, high-quality response.

 Responses from models: {{responses}}"""

+
+DEFAULT_CODE_INTERPRETER_PROMPT = """
+#### Tools Available
+
+1. **Code Interpreter**: `<code_interpreter type="code" lang="python"></code_interpreter>`
+   - You have access to a Python shell that runs directly in the user's browser, enabling fast execution of code for analysis, calculations, or problem-solving.  Use it in this response.
+   - The Python code you write can incorporate a wide array of libraries, handle data manipulation or visualization, perform API calls for web-related tasks, or tackle virtually any computational challenge. Use this flexibility to **think outside the box, craft elegant solutions, and harness Python's full potential**.
+   - To use it, **you must enclose your code within `<code_interpreter type="code" lang="python">` XML tags** and stop right away. If you don't, the code won't execute. Do NOT use triple backticks.
+   - When coding, **always aim to print meaningful outputs** (e.g., results, tables, summaries, or visuals) to better interpret and verify the findings. Avoid relying on implicit outputs; prioritize explicit and clear print statements so the results are effectively communicated to the user.  
+   - After obtaining the printed output, **always provide a concise analysis, interpretation, or next steps to help the user understand the findings or refine the outcome further.**  
+   - If the results are unclear, unexpected, or require validation, refine the code and execute it again as needed. Always aim to deliver meaningful insights from the results, iterating if necessary.  
+   - If a link is provided for an image, audio, or any file, include it in the response exactly as given to ensure the user has access to the original resource.  
+   - All responses should be communicated in the chat's primary language, ensuring seamless understanding. If the chat is multilingual, default to English for clarity.
+   - **If a link to an image, audio, or any file is provided in markdown format, explicitly display it as part of the response to ensure the user can access it easily, do NOT change the link.**
+
+Ensure that the tools are effectively utilized to achieve the highest-quality analysis for the user."""
+
+
 ####################################
 # Vector Database
 ####################################
@@ -1193,6 +1370,8 @@ CHROMA_HTTP_SSL = os.environ.get("CHROMA_HTTP_SSL", "false").lower() == "true"
 # Milvus

 MILVUS_URI = os.environ.get("MILVUS_URI", f"{DATA_DIR}/vector_db/milvus.db")
+MILVUS_DB = os.environ.get("MILVUS_DB", "default")
+MILVUS_TOKEN = os.environ.get("MILVUS_TOKEN", None)

 # Qdrant
 QDRANT_URI = os.environ.get("QDRANT_URI", None)
@@ -1211,6 +1390,9 @@ if VECTOR_DB == "pgvector" and not PGVECTOR_DB_URL.startswith("postgres"):
    raise ValueError(
        "Pgvector requires setting PGVECTOR_DB_URL or using Postgres with vector extension as the primary database."
    )
+PGVECTOR_INITIALIZE_MAX_VECTOR_LENGTH = int(
+    os.environ.get("PGVECTOR_INITIALIZE_MAX_VECTOR_LENGTH", "1536")
+)

 ####################################
 # Information Retrieval (RAG)
@@ -1570,6 +1752,11 @@ BING_SEARCH_V7_SUBSCRIPTION_KEY = PersistentConfig(
    os.environ.get("BING_SEARCH_V7_SUBSCRIPTION_KEY", ""),
 )

+EXA_API_KEY = PersistentConfig(
+    "EXA_API_KEY",
+    "rag.web.search.exa_api_key",
+    os.getenv("EXA_API_KEY", ""),
+)

 RAG_WEB_SEARCH_RESULT_COUNT = PersistentConfig(
    "RAG_WEB_SEARCH_RESULT_COUNT",
@@ -1599,6 +1786,13 @@ ENABLE_IMAGE_GENERATION = PersistentConfig(
    "image_generation.enable",
    os.environ.get("ENABLE_IMAGE_GENERATION", "").lower() == "true",
 )
+
+ENABLE_IMAGE_PROMPT_GENERATION = PersistentConfig(
+    "ENABLE_IMAGE_PROMPT_GENERATION",
+    "image_generation.prompt.enable",
+    os.environ.get("ENABLE_IMAGE_PROMPT_GENERATION", "true").lower() == "true",
+)
+
 AUTOMATIC1111_BASE_URL = PersistentConfig(
    "AUTOMATIC1111_BASE_URL",
    "image_generation.automatic1111.base_url",
@@ -1928,6 +2122,12 @@ LDAP_SERVER_PORT = PersistentConfig(
    int(os.environ.get("LDAP_SERVER_PORT", "389")),
 )

+LDAP_ATTRIBUTE_FOR_MAIL = PersistentConfig(
+    "LDAP_ATTRIBUTE_FOR_MAIL",
+    "ldap.server.attribute_for_mail",
+    os.environ.get("LDAP_ATTRIBUTE_FOR_MAIL", "mail"),
+)
+
 LDAP_ATTRIBUTE_FOR_USERNAME = PersistentConfig(
    "LDAP_ATTRIBUTE_FOR_USERNAME",
    "ldap.server.attribute_for_username",
@@ -57,7 +57,7 @@ class ERROR_MESSAGES(str, Enum):
    )

    FILE_NOT_SENT = "FILE_NOT_SENT"
-    FILE_NOT_SUPPORTED = "Oops! It seems like the file format you're trying to upload is not supported. Please upload a file with a supported format (e.g., JPG, PNG, PDF, TXT) and try again."
+    FILE_NOT_SUPPORTED = "Oops! It seems like the file format you're trying to upload is not supported. Please upload a file with a supported format and try again."

    NOT_FOUND = "We could not find what you're looking for :/"
    USER_NOT_FOUND = "We could not find what you're looking for :/"
@@ -113,6 +113,7 @@ class TASKS(str, Enum):
    TAGS_GENERATION = "tags_generation"
    EMOJI_GENERATION = "emoji_generation"
    QUERY_GENERATION = "query_generation"
+    IMAGE_PROMPT_GENERATION = "image_prompt_generation"
    AUTOCOMPLETE_GENERATION = "autocomplete_generation"
    FUNCTION_CALLING = "function_calling"
    MOA_RESPONSE_GENERATION = "moa_response_generation"
@@ -53,6 +53,13 @@ if USE_CUDA.lower() == "true":
 else:
    DEVICE_TYPE = "cpu"

+try:
+    import torch
+
+    if torch.backends.mps.is_available() and torch.backends.mps.is_built():
+        DEVICE_TYPE = "mps"
+except Exception:
+    pass

 ####################################
 # LOGGING
@@ -267,6 +274,8 @@ DATABASE_URL = os.environ.get("DATABASE_URL", f"sqlite:///{DATA_DIR}/webui.db")
 if "postgres://" in DATABASE_URL:
    DATABASE_URL = DATABASE_URL.replace("postgres://", "postgresql://")

+DATABASE_SCHEMA = os.environ.get("DATABASE_SCHEMA", None)
+
 DATABASE_POOL_SIZE = os.environ.get("DATABASE_POOL_SIZE", 0)

 if DATABASE_POOL_SIZE == "":
@@ -311,6 +320,11 @@ RESET_CONFIG_ON_START = (
    os.environ.get("RESET_CONFIG_ON_START", "False").lower() == "true"
 )

+
+ENABLE_REALTIME_CHAT_SAVE = (
+    os.environ.get("ENABLE_REALTIME_CHAT_SAVE", "False").lower() == "true"
+)
+
 ####################################
 # REDIS
 ####################################
@@ -342,14 +356,22 @@ WEBUI_SECRET_KEY = os.environ.get(
    ),  # 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_SAME_SITE = os.environ.get("WEBUI_SESSION_COOKIE_SAME_SITE", "lax")
+
+WEBUI_SESSION_COOKIE_SECURE = (
+    os.environ.get("WEBUI_SESSION_COOKIE_SECURE", "false").lower() == "true"
 )

-WEBUI_SESSION_COOKIE_SECURE = os.environ.get(
-    "WEBUI_SESSION_COOKIE_SECURE",
-    os.environ.get("WEBUI_SESSION_COOKIE_SECURE", "false").lower() == "true",
+WEBUI_AUTH_COOKIE_SAME_SITE = os.environ.get(
+    "WEBUI_AUTH_COOKIE_SAME_SITE", WEBUI_SESSION_COOKIE_SAME_SITE
+)
+
+WEBUI_AUTH_COOKIE_SECURE = (
+    os.environ.get(
+        "WEBUI_AUTH_COOKIE_SECURE",
+        os.environ.get("WEBUI_SESSION_COOKIE_SECURE", "false"),
+    ).lower()
+    == "true"
 )

 if WEBUI_AUTH and WEBUI_SECRET_KEY == "":
@@ -392,3 +414,6 @@ else:
 ####################################

 OFFLINE_MODE = os.environ.get("OFFLINE_MODE", "false").lower() == "true"
+
+if OFFLINE_MODE:
+    os.environ["HF_HUB_OFFLINE"] = "1"
@@ -7,6 +7,7 @@ from open_webui.internal.wrappers import register_connection
 from open_webui.env import (
    OPEN_WEBUI_DIR,
    DATABASE_URL,
+    DATABASE_SCHEMA,
    SRC_LOG_LEVELS,
    DATABASE_POOL_MAX_OVERFLOW,
    DATABASE_POOL_RECYCLE,
@@ -14,7 +15,7 @@ from open_webui.env import (
    DATABASE_POOL_TIMEOUT,
 )
 from peewee_migrate import Router
-from sqlalchemy import Dialect, create_engine, types
+from sqlalchemy import Dialect, create_engine, MetaData, types
 from sqlalchemy.ext.declarative import declarative_base
 from sqlalchemy.orm import scoped_session, sessionmaker
 from sqlalchemy.pool import QueuePool, NullPool
@@ -99,7 +100,8 @@ else:
 SessionLocal = sessionmaker(
    autocommit=False, autoflush=False, bind=engine, expire_on_commit=False
 )
-Base = declarative_base()
+metadata_obj = MetaData(schema=DATABASE_SCHEMA)
+Base = declarative_base(metadata=metadata_obj)
 Session = scoped_session(SessionLocal)


@@ -108,6 +108,7 @@ from open_webui.config import (
    COMFYUI_WORKFLOW,
    COMFYUI_WORKFLOW_NODES,
    ENABLE_IMAGE_GENERATION,
+    ENABLE_IMAGE_PROMPT_GENERATION,
    IMAGE_GENERATION_ENGINE,
    IMAGE_GENERATION_MODEL,
    IMAGE_SIZE,
@@ -176,6 +177,7 @@ from open_webui.config import (
    BING_SEARCH_V7_ENDPOINT,
    BING_SEARCH_V7_SUBSCRIPTION_KEY,
    BRAVE_SEARCH_API_KEY,
+    EXA_API_KEY,
    KAGI_SEARCH_API_KEY,
    MOJEEK_SEARCH_API_KEY,
    GOOGLE_PSE_API_KEY,
@@ -225,6 +227,7 @@ from open_webui.config import (
    LDAP_SERVER_LABEL,
    LDAP_SERVER_HOST,
    LDAP_SERVER_PORT,
+    LDAP_ATTRIBUTE_FOR_MAIL,
    LDAP_ATTRIBUTE_FOR_USERNAME,
    LDAP_SEARCH_FILTERS,
    LDAP_SEARCH_BASE,
@@ -254,6 +257,7 @@ from open_webui.config import (
    ENABLE_AUTOCOMPLETE_GENERATION,
    TITLE_GENERATION_PROMPT_TEMPLATE,
    TAGS_GENERATION_PROMPT_TEMPLATE,
+    IMAGE_PROMPT_GENERATION_PROMPT_TEMPLATE,
    TOOLS_FUNCTION_CALLING_PROMPT_TEMPLATE,
    QUERY_GENERATION_PROMPT_TEMPLATE,
    AUTOCOMPLETE_GENERATION_PROMPT_TEMPLATE,
@@ -437,6 +441,7 @@ app.state.config.ENABLE_LDAP = ENABLE_LDAP
 app.state.config.LDAP_SERVER_LABEL = LDAP_SERVER_LABEL
 app.state.config.LDAP_SERVER_HOST = LDAP_SERVER_HOST
 app.state.config.LDAP_SERVER_PORT = LDAP_SERVER_PORT
+app.state.config.LDAP_ATTRIBUTE_FOR_MAIL = LDAP_ATTRIBUTE_FOR_MAIL
 app.state.config.LDAP_ATTRIBUTE_FOR_USERNAME = LDAP_ATTRIBUTE_FOR_USERNAME
 app.state.config.LDAP_APP_DN = LDAP_APP_DN
 app.state.config.LDAP_APP_PASSWORD = LDAP_APP_PASSWORD
@@ -519,6 +524,7 @@ app.state.config.SEARCHAPI_ENGINE = SEARCHAPI_ENGINE
 app.state.config.JINA_API_KEY = JINA_API_KEY
 app.state.config.BING_SEARCH_V7_ENDPOINT = BING_SEARCH_V7_ENDPOINT
 app.state.config.BING_SEARCH_V7_SUBSCRIPTION_KEY = BING_SEARCH_V7_SUBSCRIPTION_KEY
+app.state.config.EXA_API_KEY = EXA_API_KEY

 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
@@ -572,6 +578,7 @@ app.state.EMBEDDING_FUNCTION = get_embedding_function(

 app.state.config.IMAGE_GENERATION_ENGINE = IMAGE_GENERATION_ENGINE
 app.state.config.ENABLE_IMAGE_GENERATION = ENABLE_IMAGE_GENERATION
+app.state.config.ENABLE_IMAGE_PROMPT_GENERATION = ENABLE_IMAGE_PROMPT_GENERATION

 app.state.config.IMAGES_OPENAI_API_BASE_URL = IMAGES_OPENAI_API_BASE_URL
 app.state.config.IMAGES_OPENAI_API_KEY = IMAGES_OPENAI_API_KEY
@@ -642,6 +649,10 @@ app.state.config.ENABLE_TAGS_GENERATION = ENABLE_TAGS_GENERATION

 app.state.config.TITLE_GENERATION_PROMPT_TEMPLATE = TITLE_GENERATION_PROMPT_TEMPLATE
 app.state.config.TAGS_GENERATION_PROMPT_TEMPLATE = TAGS_GENERATION_PROMPT_TEMPLATE
+app.state.config.IMAGE_PROMPT_GENERATION_PROMPT_TEMPLATE = (
+    IMAGE_PROMPT_GENERATION_PROMPT_TEMPLATE
+)
+
 app.state.config.TOOLS_FUNCTION_CALLING_PROMPT_TEMPLATE = (
    TOOLS_FUNCTION_CALLING_PROMPT_TEMPLATE
 )
@@ -850,6 +861,7 @@ async def chat_completion(
        if model_id not in request.app.state.MODELS:
            raise Exception("Model not found")
        model = request.app.state.MODELS[model_id]
+        model_info = Models.get_model_by_id(model_id)

        # Check if user has access to the model
        if not BYPASS_MODEL_ACCESS_CONTROL and user.role == "user":
@@ -866,12 +878,25 @@ async def chat_completion(
            "tool_ids": form_data.get("tool_ids", None),
            "files": form_data.get("files", None),
            "features": form_data.get("features", None),
+            "variables": form_data.get("variables", None),
+            "model": model_info,
+            **(
+                {"function_calling": "native"}
+                if form_data.get("params", {}).get("function_calling") == "native"
+                or (
+                    model_info
+                    and model_info.params.model_dump().get("function_calling")
+                    == "native"
+                )
+                else {}
+            ),
        }
        form_data["metadata"] = metadata

-        form_data, events = await process_chat_payload(
+        form_data, metadata, events = await process_chat_payload(
            request, form_data, metadata, user, model
        )
+
    except Exception as e:
        raise HTTPException(
            status_code=status.HTTP_400_BAD_REQUEST,
@@ -880,6 +905,7 @@ async def chat_completion(

    try:
        response = await chat_completion_handler(request, form_data, user)
+
        return await process_chat_response(
            request, response, form_data, user, events, metadata, tasks
        )
@@ -998,10 +1024,6 @@ async def get_app_config(request: Request):
                else {}
            ),
        },
-        "google_drive": {
-            "client_id": GOOGLE_DRIVE_CLIENT_ID.value,
-            "api_key": GOOGLE_DRIVE_API_KEY.value,
-        },
        **(
            {
                "default_models": app.state.config.DEFAULT_MODELS,
@@ -1021,6 +1043,10 @@ async def get_app_config(request: Request):
                    "max_count": app.state.config.FILE_MAX_COUNT,
                },
                "permissions": {**app.state.config.USER_PERMISSIONS},
+                "google_drive": {
+                    "client_id": GOOGLE_DRIVE_CLIENT_ID.value,
+                    "api_key": GOOGLE_DRIVE_API_KEY.value,
+                },
            }
            if user is not None
            else {}
@@ -1054,7 +1080,7 @@ async def get_app_version():


@app.get("/api/version/updates")
-async def get_app_latest_release_version():
+async def get_app_latest_release_version(user=Depends(get_verified_user)):
    if OFFLINE_MODE:
        log.debug(
            f"Offline mode is enabled, returning current version as latest version"
@@ -0,0 +1,70 @@
+"""Update message & channel tables
+
+Revision ID: 3781e22d8b01
+Revises: 7826ab40b532
+Create Date: 2024-12-30 03:00:00.000000
+
+"""
+
+from alembic import op
+import sqlalchemy as sa
+
+revision = "3781e22d8b01"
+down_revision = "7826ab40b532"
+branch_labels = None
+depends_on = None
+
+
+def upgrade():
+    # Add 'type' column to the 'channel' table
+    op.add_column(
+        "channel",
+        sa.Column(
+            "type",
+            sa.Text(),
+            nullable=True,
+        ),
+    )
+
+    # Add 'parent_id' column to the 'message' table for threads
+    op.add_column(
+        "message",
+        sa.Column("parent_id", sa.Text(), nullable=True),
+    )
+
+    op.create_table(
+        "message_reaction",
+        sa.Column(
+            "id", sa.Text(), nullable=False, primary_key=True, unique=True
+        ),  # Unique reaction ID
+        sa.Column("user_id", sa.Text(), nullable=False),  # User who reacted
+        sa.Column(
+            "message_id", sa.Text(), nullable=False
+        ),  # Message that was reacted to
+        sa.Column(
+            "name", sa.Text(), nullable=False
+        ),  # Reaction name (e.g. "thumbs_up")
+        sa.Column(
+            "created_at", sa.BigInteger(), nullable=True
+        ),  # Timestamp of when the reaction was added
+    )
+
+    op.create_table(
+        "channel_member",
+        sa.Column(
+            "id", sa.Text(), nullable=False, primary_key=True, unique=True
+        ),  # Record ID for the membership row
+        sa.Column("channel_id", sa.Text(), nullable=False),  # Associated channel
+        sa.Column("user_id", sa.Text(), nullable=False),  # Associated user
+        sa.Column(
+            "created_at", sa.BigInteger(), nullable=True
+        ),  # Timestamp of when the user joined the channel
+    )
+
+
+def downgrade():
+    # Revert 'type' column addition to the 'channel' table
+    op.drop_column("channel", "type")
+    op.drop_column("message", "parent_id")
+    op.drop_table("message_reaction")
+    op.drop_table("channel_member")
@@ -21,6 +21,7 @@ class Channel(Base):

    id = Column(Text, primary_key=True)
    user_id = Column(Text)
+    type = Column(Text, nullable=True)

    name = Column(Text)
    description = Column(Text, nullable=True)
@@ -38,9 +39,11 @@ class ChannelModel(BaseModel):

    id: str
    user_id: str
-    description: Optional[str] = None
+    type: Optional[str] = None

    name: str
+    description: Optional[str] = None
+
    data: Optional[dict] = None
    meta: Optional[dict] = None
    access_control: Optional[dict] = None
@@ -64,12 +67,13 @@ class ChannelForm(BaseModel):

 class ChannelTable:
    def insert_new_channel(
-        self, form_data: ChannelForm, user_id: str
+        self, type: Optional[str], form_data: ChannelForm, user_id: str
    ) -> Optional[ChannelModel]:
        with get_db() as db:
            channel = ChannelModel(
                **{
                    **form_data.model_dump(),
+                    "type": type,
                    "name": form_data.name.lower(),
                    "id": str(uuid.uuid4()),
                    "user_id": user_id,
@@ -212,6 +212,15 @@ class ChatTable:

        return chat.chat.get("history", {}).get("messages", {}) or {}

+    def get_message_by_id_and_message_id(
+        self, id: str, message_id: str
+    ) -> Optional[dict]:
+        chat = self.get_chat_by_id(id)
+        if chat is None:
+            return None
+
+        return chat.chat.get("history", {}).get("messages", {}).get(message_id, {})
+
    def upsert_message_to_chat_by_id_and_message_id(
        self, id: str, message_id: str, message: dict
    ) -> Optional[ChatModel]:
@@ -384,7 +393,7 @@ class ChatTable:
        limit: int = 50,
    ) -> list[ChatModel]:
        with get_db() as db:
-            query = db.query(Chat).filter_by(user_id=user_id).filter_by(folder_id=None)
+            query = db.query(Chat).filter_by(user_id=user_id)
            if not include_archived:
                query = query.filter_by(archived=False)

@@ -460,6 +469,8 @@ class ChatTable:
    def get_chat_by_share_id(self, id: str) -> Optional[ChatModel]:
        try:
            with get_db() as db:
+                # it is possible that the shared link was deleted. hence,
+                # we check if the chat is still shared by checkng if a chat with the share_id exists
                chat = db.query(Chat).filter_by(share_id=id).first()

                if chat:
@@ -80,12 +80,11 @@ class GroupResponse(BaseModel):
 class GroupForm(BaseModel):
    name: str
    description: str
+    permissions: Optional[dict] = None


 class GroupUpdateForm(GroupForm):
-    permissions: Optional[dict] = None
    user_ids: Optional[list[str]] = None
-    admin_ids: Optional[list[str]] = None


 class GroupTable:
@@ -95,7 +94,7 @@ class GroupTable:
        with get_db() as db:
            group = GroupModel(
                **{
-                    **form_data.model_dump(),
+                    **form_data.model_dump(exclude_none=True),
                    "id": str(uuid.uuid4()),
                    "user_id": user_id,
                    "created_at": int(time.time()),
@@ -189,5 +188,24 @@ class GroupTable:
            except Exception:
                return False

+    def remove_user_from_all_groups(self, user_id: str) -> bool:
+        with get_db() as db:
+            try:
+                groups = self.get_groups_by_member_id(user_id)
+
+                for group in groups:
+                    group.user_ids.remove(user_id)
+                    db.query(Group).filter_by(id=group.id).update(
+                        {
+                            "user_ids": group.user_ids,
+                            "updated_at": int(time.time()),
+                        }
+                    )
+                    db.commit()
+
+                return True
+            except Exception:
+                return False
+

 Groups = GroupTable()
@@ -17,6 +17,25 @@ from sqlalchemy.sql import exists
 ####################


+class MessageReaction(Base):
+    __tablename__ = "message_reaction"
+    id = Column(Text, primary_key=True)
+    user_id = Column(Text)
+    message_id = Column(Text)
+    name = Column(Text)
+    created_at = Column(BigInteger)
+
+
+class MessageReactionModel(BaseModel):
+    model_config = ConfigDict(from_attributes=True)
+
+    id: str
+    user_id: str
+    message_id: str
+    name: str
+    created_at: int  # timestamp in epoch
+
+
 class Message(Base):
    __tablename__ = "message"
    id = Column(Text, primary_key=True)
@@ -24,6 +43,8 @@ class Message(Base):
    user_id = Column(Text)
    channel_id = Column(Text, nullable=True)

+    parent_id = Column(Text, nullable=True)
+
    content = Column(Text)
    data = Column(JSON, nullable=True)
    meta = Column(JSON, nullable=True)
@@ -39,6 +60,8 @@ class MessageModel(BaseModel):
    user_id: str
    channel_id: Optional[str] = None

+    parent_id: Optional[str] = None
+
    content: str
    data: Optional[dict] = None
    meta: Optional[dict] = None
@@ -54,10 +77,23 @@ class MessageModel(BaseModel):

 class MessageForm(BaseModel):
    content: str
+    parent_id: Optional[str] = None
    data: Optional[dict] = None
    meta: Optional[dict] = None


+class Reactions(BaseModel):
+    name: str
+    user_ids: list[str]
+    count: int
+
+
+class MessageResponse(MessageModel):
+    latest_reply_at: Optional[int]
+    reply_count: int
+    reactions: list[Reactions]
+
+
 class MessageTable:
    def insert_new_message(
        self, form_data: MessageForm, channel_id: str, user_id: str
@@ -71,6 +107,7 @@ class MessageTable:
                    "id": id,
                    "user_id": user_id,
                    "channel_id": channel_id,
+                    "parent_id": form_data.parent_id,
                    "content": form_data.content,
                    "data": form_data.data,
                    "meta": form_data.meta,
@@ -85,10 +122,40 @@ class MessageTable:
            db.refresh(result)
            return MessageModel.model_validate(result) if result else None

-    def get_message_by_id(self, id: str) -> Optional[MessageModel]:
+    def get_message_by_id(self, id: str) -> Optional[MessageResponse]:
        with get_db() as db:
            message = db.get(Message, id)
-            return MessageModel.model_validate(message) if message else None
+            if not message:
+                return None
+
+            reactions = self.get_reactions_by_message_id(id)
+            replies = self.get_replies_by_message_id(id)
+
+            return MessageResponse(
+                **{
+                    **MessageModel.model_validate(message).model_dump(),
+                    "latest_reply_at": replies[0].created_at if replies else None,
+                    "reply_count": len(replies),
+                    "reactions": reactions,
+                }
+            )
+
+    def get_replies_by_message_id(self, id: str) -> list[MessageModel]:
+        with get_db() as db:
+            all_messages = (
+                db.query(Message)
+                .filter_by(parent_id=id)
+                .order_by(Message.created_at.desc())
+                .all()
+            )
+            return [MessageModel.model_validate(message) for message in all_messages]
+
+    def get_reply_user_ids_by_message_id(self, id: str) -> list[str]:
+        with get_db() as db:
+            return [
+                message.user_id
+                for message in db.query(Message).filter_by(parent_id=id).all()
+            ]

    def get_messages_by_channel_id(
        self, channel_id: str, skip: int = 0, limit: int = 50
@@ -96,7 +163,7 @@ class MessageTable:
        with get_db() as db:
            all_messages = (
                db.query(Message)
-                .filter_by(channel_id=channel_id)
+                .filter_by(channel_id=channel_id, parent_id=None)
                .order_by(Message.created_at.desc())
                .offset(skip)
                .limit(limit)
@@ -104,18 +171,28 @@ class MessageTable:
            )
            return [MessageModel.model_validate(message) for message in all_messages]

-    def get_messages_by_user_id(
-        self, user_id: str, skip: int = 0, limit: int = 50
+    def get_messages_by_parent_id(
+        self, channel_id: str, parent_id: str, skip: int = 0, limit: int = 50
    ) -> list[MessageModel]:
        with get_db() as db:
+            message = db.get(Message, parent_id)
+
+            if not message:
+                return []
+
            all_messages = (
                db.query(Message)
-                .filter_by(user_id=user_id)
+                .filter_by(channel_id=channel_id, parent_id=parent_id)
                .order_by(Message.created_at.desc())
                .offset(skip)
                .limit(limit)
                .all()
            )
+
+            # If length of all_messages is less than limit, then add the parent message
+            if len(all_messages) < limit:
+                all_messages.append(message)
+
            return [MessageModel.model_validate(message) for message in all_messages]

    def update_message_by_id(
@@ -131,9 +208,70 @@ class MessageTable:
            db.refresh(message)
            return MessageModel.model_validate(message) if message else None

+    def add_reaction_to_message(
+        self, id: str, user_id: str, name: str
+    ) -> Optional[MessageReactionModel]:
+        with get_db() as db:
+            reaction_id = str(uuid.uuid4())
+            reaction = MessageReactionModel(
+                id=reaction_id,
+                user_id=user_id,
+                message_id=id,
+                name=name,
+                created_at=int(time.time_ns()),
+            )
+            result = MessageReaction(**reaction.model_dump())
+            db.add(result)
+            db.commit()
+            db.refresh(result)
+            return MessageReactionModel.model_validate(result) if result else None
+
+    def get_reactions_by_message_id(self, id: str) -> list[Reactions]:
+        with get_db() as db:
+            all_reactions = db.query(MessageReaction).filter_by(message_id=id).all()
+
+            reactions = {}
+            for reaction in all_reactions:
+                if reaction.name not in reactions:
+                    reactions[reaction.name] = {
+                        "name": reaction.name,
+                        "user_ids": [],
+                        "count": 0,
+                    }
+                reactions[reaction.name]["user_ids"].append(reaction.user_id)
+                reactions[reaction.name]["count"] += 1
+
+            return [Reactions(**reaction) for reaction in reactions.values()]
+
+    def remove_reaction_by_id_and_user_id_and_name(
+        self, id: str, user_id: str, name: str
+    ) -> bool:
+        with get_db() as db:
+            db.query(MessageReaction).filter_by(
+                message_id=id, user_id=user_id, name=name
+            ).delete()
+            db.commit()
+            return True
+
+    def delete_reactions_by_id(self, id: str) -> bool:
+        with get_db() as db:
+            db.query(MessageReaction).filter_by(message_id=id).delete()
+            db.commit()
+            return True
+
+    def delete_replies_by_id(self, id: str) -> bool:
+        with get_db() as db:
+            db.query(Message).filter_by(parent_id=id).delete()
+            db.commit()
+            return True
+
    def delete_message_by_id(self, id: str) -> bool:
        with get_db() as db:
            db.query(Message).filter_by(id=id).delete()
+
+            # Delete all reactions to this message
+            db.query(MessageReaction).filter_by(message_id=id).delete()
+
            db.commit()
            return True

@@ -2,7 +2,12 @@ import time
 from typing import Optional

 from open_webui.internal.db import Base, JSONField, get_db
+
+
 from open_webui.models.chats import Chats
+from open_webui.models.groups import Groups
+
+
 from pydantic import BaseModel, ConfigDict
 from sqlalchemy import BigInteger, Column, String, Text

@@ -268,9 +273,11 @@ class UsersTable:

    def delete_user_by_id(self, id: str) -> bool:
        try:
+            # Remove User from Groups
+            Groups.remove_user_from_all_groups(id)
+
            # Delete User Chats
            result = Chats.delete_chats_by_user_id(id)
-
            if result:
                with get_db() as db:
                    # Delete User
@@ -300,5 +307,10 @@ class UsersTable:
        except Exception:
            return None

+    def get_valid_user_ids(self, user_ids: list[str]) -> list[str]:
+        with get_db() as db:
+            users = db.query(User).filter(User.id.in_(user_ids)).all()
+            return [user.id for user in users]
+

 Users = UsersTable()
@@ -11,10 +11,17 @@ from langchain.retrievers import ContextualCompressionRetriever, EnsembleRetriev
 from langchain_community.retrievers import BM25Retriever
 from langchain_core.documents import Document

+
+from open_webui.config import VECTOR_DB
 from open_webui.retrieval.vector.connector import VECTOR_DB_CLIENT
 from open_webui.utils.misc import get_last_user_message
+from open_webui.models.users import UserModel

-from open_webui.env import SRC_LOG_LEVELS
+from open_webui.env import (
+    SRC_LOG_LEVELS,
+    OFFLINE_MODE,
+    ENABLE_FORWARD_USER_INFO_HEADERS,
+)

 log = logging.getLogger(__name__)
 log.setLevel(SRC_LOG_LEVELS["RAG"])
@@ -59,9 +66,7 @@ class VectorSearchRetriever(BaseRetriever):


 def query_doc(
-    collection_name: str,
-    query_embedding: list[float],
-    k: int,
+    collection_name: str, query_embedding: list[float], k: int, user: UserModel = None
 ):
    try:
        result = VECTOR_DB_CLIENT.search(
@@ -199,7 +204,12 @@ def query_collection(
            else:
                pass

-    return merge_and_sort_query_results(results, k=k)
+    if VECTOR_DB == "chroma":
+        # Chroma uses unconventional cosine similarity, so we don't need to reverse the results
+        # https://docs.trychroma.com/docs/collections/configure#configuring-chroma-collections
+        return merge_and_sort_query_results(results, k=k, reverse=False)
+    else:
+        return merge_and_sort_query_results(results, k=k, reverse=True)


 def query_collection_with_hybrid_search(
@@ -235,7 +245,12 @@ def query_collection_with_hybrid_search(
            "Hybrid search failed for all collections. Using Non hybrid search as fallback."
        )

-    return merge_and_sort_query_results(results, k=k, reverse=True)
+    if VECTOR_DB == "chroma":
+        # Chroma uses unconventional cosine similarity, so we don't need to reverse the results
+        # https://docs.trychroma.com/docs/collections/configure#configuring-chroma-collections
+        return merge_and_sort_query_results(results, k=k, reverse=False)
+    else:
+        return merge_and_sort_query_results(results, k=k, reverse=True)


 def get_embedding_function(
@@ -247,26 +262,31 @@ def get_embedding_function(
    embedding_batch_size,
 ):
    if embedding_engine == "":
-        return lambda query: embedding_function.encode(query).tolist()
+        return lambda query, user=None: embedding_function.encode(query).tolist()
    elif embedding_engine in ["ollama", "openai"]:
-        func = lambda query: generate_embeddings(
+        func = lambda query, user=None: generate_embeddings(
            engine=embedding_engine,
            model=embedding_model,
            text=query,
            url=url,
            key=key,
+            user=user,
        )

-        def generate_multiple(query, func):
+        def generate_multiple(query, user, func):
            if isinstance(query, list):
                embeddings = []
                for i in range(0, len(query), embedding_batch_size):
-                    embeddings.extend(func(query[i : i + embedding_batch_size]))
+                    embeddings.extend(
+                        func(query[i : i + embedding_batch_size], user=user)
+                    )
                return embeddings
            else:
-                return func(query)
+                return func(query, user)

-        return lambda query: generate_multiple(query, func)
+        return lambda query, user=None: generate_multiple(query, user, func)
+    else:
+        raise ValueError(f"Unknown embedding engine: {embedding_engine}")


 def get_sources_from_files(
@@ -375,6 +395,9 @@ def get_model_path(model: str, update_model: bool = False):

    local_files_only = not update_model

+    if OFFLINE_MODE:
+        local_files_only = True
+
    snapshot_kwargs = {
        "cache_dir": cache_dir,
        "local_files_only": local_files_only,
@@ -408,7 +431,11 @@ def get_model_path(model: str, update_model: bool = False):


 def generate_openai_batch_embeddings(
-    model: str, texts: list[str], url: str = "https://api.openai.com/v1", key: str = ""
+    model: str,
+    texts: list[str],
+    url: str = "https://api.openai.com/v1",
+    key: str = "",
+    user: UserModel = None,
 ) -> Optional[list[list[float]]]:
    try:
        r = requests.post(
@@ -416,6 +443,16 @@ def generate_openai_batch_embeddings(
            headers={
                "Content-Type": "application/json",
                "Authorization": f"Bearer {key}",
+                **(
+                    {
+                        "X-OpenWebUI-User-Name": user.name,
+                        "X-OpenWebUI-User-Id": user.id,
+                        "X-OpenWebUI-User-Email": user.email,
+                        "X-OpenWebUI-User-Role": user.role,
+                    }
+                    if ENABLE_FORWARD_USER_INFO_HEADERS and user
+                    else {}
+                ),
            },
            json={"input": texts, "model": model},
        )
@@ -431,7 +468,7 @@ def generate_openai_batch_embeddings(


 def generate_ollama_batch_embeddings(
-    model: str, texts: list[str], url: str, key: str = ""
+    model: str, texts: list[str], url: str, key: str = "", user: UserModel = None
 ) -> Optional[list[list[float]]]:
    try:
        r = requests.post(
@@ -439,6 +476,16 @@ def generate_ollama_batch_embeddings(
            headers={
                "Content-Type": "application/json",
                "Authorization": f"Bearer {key}",
+                **(
+                    {
+                        "X-OpenWebUI-User-Name": user.name,
+                        "X-OpenWebUI-User-Id": user.id,
+                        "X-OpenWebUI-User-Email": user.email,
+                        "X-OpenWebUI-User-Role": user.role,
+                    }
+                    if ENABLE_FORWARD_USER_INFO_HEADERS
+                    else {}
+                ),
            },
            json={"input": texts, "model": model},
        )
@@ -457,22 +504,29 @@ def generate_ollama_batch_embeddings(
 def generate_embeddings(engine: str, model: str, text: Union[str, list[str]], **kwargs):
    url = kwargs.get("url", "")
    key = kwargs.get("key", "")
+    user = kwargs.get("user")

    if engine == "ollama":
        if isinstance(text, list):
            embeddings = generate_ollama_batch_embeddings(
-                **{"model": model, "texts": text, "url": url, "key": key}
+                **{"model": model, "texts": text, "url": url, "key": key, "user": user}
            )
        else:
            embeddings = generate_ollama_batch_embeddings(
-                **{"model": model, "texts": [text], "url": url, "key": key}
+                **{
+                    "model": model,
+                    "texts": [text],
+                    "url": url,
+                    "key": key,
+                    "user": user,
+                }
            )
        return embeddings[0] if isinstance(text, str) else embeddings
    elif engine == "openai":
        if isinstance(text, list):
-            embeddings = generate_openai_batch_embeddings(model, text, url, key)
+            embeddings = generate_openai_batch_embeddings(model, text, url, key, user)
        else:
-            embeddings = generate_openai_batch_embeddings(model, [text], url, key)
+            embeddings = generate_openai_batch_embeddings(model, [text], url, key, user)

        return embeddings[0] if isinstance(text, str) else embeddings

@@ -51,8 +51,8 @@ class ChromaClient:

    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]
+        collection_names = self.client.list_collections()
+        return collection_name in collection_names

    def delete_collection(self, collection_name: str):
        # Delete the collection based on the collection name.
@@ -7,13 +7,18 @@ from typing import Optional
 from open_webui.retrieval.vector.main import VectorItem, SearchResult, GetResult
 from open_webui.config import (
    MILVUS_URI,
+    MILVUS_DB,
+    MILVUS_TOKEN,
 )


 class MilvusClient:
    def __init__(self):
        self.collection_prefix = "open_webui"
-        self.client = Client(uri=MILVUS_URI)
+        if MILVUS_TOKEN is None:
+            self.client = Client(uri=MILVUS_URI, database=MILVUS_DB)
+        else:
+            self.client = Client(uri=MILVUS_URI, database=MILVUS_DB, token=MILVUS_TOKEN)

    def _result_to_get_result(self, result) -> GetResult:
        ids = []
@@ -5,9 +5,11 @@ from sqlalchemy import (
    create_engine,
    Column,
    Integer,
+    MetaData,
    select,
    text,
    Text,
+    Table,
    values,
 )
 from sqlalchemy.sql import true
@@ -17,11 +19,12 @@ from sqlalchemy.orm import declarative_base, scoped_session, sessionmaker
 from sqlalchemy.dialects.postgresql import JSONB, array
 from pgvector.sqlalchemy import Vector
 from sqlalchemy.ext.mutable import MutableDict
+from sqlalchemy.exc import NoSuchTableError

 from open_webui.retrieval.vector.main import VectorItem, SearchResult, GetResult
-from open_webui.config import PGVECTOR_DB_URL
+from open_webui.config import PGVECTOR_DB_URL, PGVECTOR_INITIALIZE_MAX_VECTOR_LENGTH

-VECTOR_LENGTH = 1536
+VECTOR_LENGTH = PGVECTOR_INITIALIZE_MAX_VECTOR_LENGTH
 Base = declarative_base()


@@ -56,6 +59,9 @@ class PgvectorClient:
            # Ensure the pgvector extension is available
            self.session.execute(text("CREATE EXTENSION IF NOT EXISTS vector;"))

+            # Check vector length consistency
+            self.check_vector_length()
+
            # Create the tables if they do not exist
            # Base.metadata.create_all requires a bind (engine or connection)
            # Get the connection from the session
@@ -82,6 +88,41 @@ class PgvectorClient:
            print(f"Error during initialization: {e}")
            raise

+    def check_vector_length(self) -> None:
+        """
+        Check if the VECTOR_LENGTH matches the existing vector column dimension in the database.
+        Raises an exception if there is a mismatch.
+        """
+        metadata = MetaData()
+        try:
+            # Attempt to reflect the 'document_chunk' table
+            document_chunk_table = Table(
+                "document_chunk", metadata, autoload_with=self.session.bind
+            )
+        except NoSuchTableError:
+            # Table does not exist; no action needed
+            return
+
+        # Proceed to check the vector column
+        if "vector" in document_chunk_table.columns:
+            vector_column = document_chunk_table.columns["vector"]
+            vector_type = vector_column.type
+            if isinstance(vector_type, Vector):
+                db_vector_length = vector_type.dim
+                if db_vector_length != VECTOR_LENGTH:
+                    raise Exception(
+                        f"VECTOR_LENGTH {VECTOR_LENGTH} does not match existing vector column dimension {db_vector_length}. "
+                        "Cannot change vector size after initialization without migrating the data."
+                    )
+            else:
+                raise Exception(
+                    "The 'vector' column exists but is not of type 'Vector'."
+                )
+        else:
+            raise Exception(
+                "The 'vector' column does not exist in the 'document_chunk' table."
+            )
+
    def adjust_vector_length(self, vector: List[float]) -> List[float]:
        # Adjust vector to have length VECTOR_LENGTH
        current_length = len(vector)
@@ -23,7 +23,7 @@ def search_bing(
    filter_list: Optional[list[str]] = None,
 ) -> list[SearchResult]:
    mkt = locale
-    params = {"q": query, "mkt": mkt, "answerCount": count}
+    params = {"q": query, "mkt": mkt, "count": count}
    headers = {"Ocp-Apim-Subscription-Key": subscription_key}

    try:
@@ -0,0 +1,76 @@
+import logging
+from dataclasses import dataclass
+from typing import Optional
+
+import requests
+from open_webui.env import SRC_LOG_LEVELS
+from open_webui.retrieval.web.main import SearchResult
+
+log = logging.getLogger(__name__)
+log.setLevel(SRC_LOG_LEVELS["RAG"])
+
+EXA_API_BASE = "https://api.exa.ai"
+
+
+@dataclass
+class ExaResult:
+    url: str
+    title: str
+    text: str
+
+
+def search_exa(
+    api_key: str,
+    query: str,
+    count: int,
+    filter_list: Optional[list[str]] = None,
+) -> list[SearchResult]:
+    """Search using Exa Search API and return the results as a list of SearchResult objects.
+
+    Args:
+        api_key (str): A Exa Search API key
+        query (str): The query to search for
+        count (int): Number of results to return
+        filter_list (Optional[list[str]]): List of domains to filter results by
+    """
+    log.info(f"Searching with Exa for query: {query}")
+
+    headers = {"Authorization": f"Bearer {api_key}", "Content-Type": "application/json"}
+
+    payload = {
+        "query": query,
+        "numResults": count or 5,
+        "includeDomains": filter_list,
+        "contents": {"text": True, "highlights": True},
+        "type": "auto",  # Use the auto search type (keyword or neural)
+    }
+
+    try:
+        response = requests.post(
+            f"{EXA_API_BASE}/search", headers=headers, json=payload
+        )
+        response.raise_for_status()
+        data = response.json()
+
+        results = []
+        for result in data["results"]:
+            results.append(
+                ExaResult(
+                    url=result["url"],
+                    title=result["title"],
+                    text=result["text"],
+                )
+            )
+
+        log.info(f"Found {len(results)} results")
+        return [
+            SearchResult(
+                link=result.url,
+                title=result.title,
+                snippet=result.text,
+            )
+            for result in results
+        ]
+    except Exception as e:
+        log.error(f"Error searching Exa: {e}")
+        return []
@@ -1,3 +1,5 @@
+import validators
+
 from typing import Optional
 from urllib.parse import urlparse

@@ -10,6 +12,8 @@ def get_filtered_results(results, filter_list):
    filtered_results = []
    for result in results:
        url = result.get("url") or result.get("link", "")
+        if not validators.url(url):
+            continue
        domain = urlparse(url).netloc
        if any(domain.endswith(filtered_domain) for filtered_domain in filter_list):
            filtered_results.append(result)
@@ -683,7 +683,7 @@
 				"age": "October 29, 2022",
 				"extra_snippets": [
 					"You can pass many options to the configure script; run ./configure --help to find out more. On macOS case-insensitive file systems and on Cygwin, the executable is called python.exe; elsewhere it's just python.",
-					"Building a complete Python installation requires the use of various additional third-party libraries, depending on your build platform and configure options. Not all standard library modules are buildable or useable on all platforms. Refer to the Install dependencies section of the Developer Guide for current detailed information on dependencies for various Linux distributions and macOS.",
+					"Building a complete Python installation requires the use of various additional third-party libraries, depending on your build platform and configure options. Not all standard library modules are buildable or usable on all platforms. Refer to the Install dependencies section of the Developer Guide for current detailed information on dependencies for various Linux distributions and macOS.",
 					"To get an optimized build of Python, configure --enable-optimizations before you run make. This sets the default make targets up to enable Profile Guided Optimization (PGO) and may be used to auto-enable Link Time Optimization (LTO) on some platforms. For more details, see the sections below.",
 					"Copyright © 2001-2024 Python Software Foundation. All rights reserved."
 				]
@@ -43,6 +43,17 @@ def validate_url(url: Union[str, Sequence[str]]):
        return False


+def safe_validate_urls(url: Sequence[str]) -> Sequence[str]:
+    valid_urls = []
+    for u in url:
+        try:
+            if validate_url(u):
+                valid_urls.append(u)
+        except ValueError:
+            continue
+    return valid_urls
+
+
 def resolve_hostname(hostname):
    # Get address information
    addr_info = socket.getaddrinfo(hostname, None)
@@ -86,11 +97,11 @@ def get_web_loader(
    verify_ssl: bool = True,
    requests_per_second: int = 2,
 ):
-    # Check if the URL is valid
-    if not validate_url(urls):
-        raise ValueError(ERROR_MESSAGES.INVALID_URL)
+    # Check if the URLs are valid
+    safe_urls = safe_validate_urls([urls] if isinstance(urls, str) else urls)
+
    return SafeWebBaseLoader(
-        urls,
+        safe_urls,
        verify_ssl=verify_ssl,
        requests_per_second=requests_per_second,
        continue_on_failure=True,
@@ -25,8 +25,8 @@ from open_webui.env import (
    WEBUI_AUTH,
    WEBUI_AUTH_TRUSTED_EMAIL_HEADER,
    WEBUI_AUTH_TRUSTED_NAME_HEADER,
-    WEBUI_SESSION_COOKIE_SAME_SITE,
-    WEBUI_SESSION_COOKIE_SECURE,
+    WEBUI_AUTH_COOKIE_SAME_SITE,
+    WEBUI_AUTH_COOKIE_SECURE,
    SRC_LOG_LEVELS,
 )
 from fastapi import APIRouter, Depends, HTTPException, Request, status
@@ -51,7 +51,7 @@ from open_webui.utils.access_control import get_permissions
 from typing import Optional, List

 from ssl import CERT_REQUIRED, PROTOCOL_TLS
-from ldap3 import Server, Connection, ALL, Tls
+from ldap3 import Server, Connection, NONE, Tls
 from ldap3.utils.conv import escape_filter_chars

 router = APIRouter()
@@ -95,8 +95,8 @@ async def get_session_user(
        value=token,
        expires=datetime_expires_at,
        httponly=True,  # Ensures the cookie is not accessible via JavaScript
-        samesite=WEBUI_SESSION_COOKIE_SAME_SITE,
-        secure=WEBUI_SESSION_COOKIE_SECURE,
+        samesite=WEBUI_AUTH_COOKIE_SAME_SITE,
+        secure=WEBUI_AUTH_COOKIE_SECURE,
    )

    user_permissions = get_permissions(
@@ -164,12 +164,13 @@ async def update_password(
 ############################
 # LDAP Authentication
 ############################
-@router.post("/ldap", response_model=SigninResponse)
+@router.post("/ldap", response_model=SessionUserResponse)
 async def ldap_auth(request: Request, response: Response, form_data: LdapForm):
    ENABLE_LDAP = request.app.state.config.ENABLE_LDAP
    LDAP_SERVER_LABEL = request.app.state.config.LDAP_SERVER_LABEL
    LDAP_SERVER_HOST = request.app.state.config.LDAP_SERVER_HOST
    LDAP_SERVER_PORT = request.app.state.config.LDAP_SERVER_PORT
+    LDAP_ATTRIBUTE_FOR_MAIL = request.app.state.config.LDAP_ATTRIBUTE_FOR_MAIL
    LDAP_ATTRIBUTE_FOR_USERNAME = request.app.state.config.LDAP_ATTRIBUTE_FOR_USERNAME
    LDAP_SEARCH_BASE = request.app.state.config.LDAP_SEARCH_BASE
    LDAP_SEARCH_FILTERS = request.app.state.config.LDAP_SEARCH_FILTERS
@@ -201,7 +202,7 @@ async def ldap_auth(request: Request, response: Response, form_data: LdapForm):
        server = Server(
            host=LDAP_SERVER_HOST,
            port=LDAP_SERVER_PORT,
-            get_info=ALL,
+            get_info=NONE,
            use_ssl=LDAP_USE_TLS,
            tls=tls,
        )
@@ -218,7 +219,11 @@ async def ldap_auth(request: Request, response: Response, form_data: LdapForm):
        search_success = connection_app.search(
            search_base=LDAP_SEARCH_BASE,
            search_filter=f"(&({LDAP_ATTRIBUTE_FOR_USERNAME}={escape_filter_chars(form_data.user.lower())}){LDAP_SEARCH_FILTERS})",
-            attributes=[f"{LDAP_ATTRIBUTE_FOR_USERNAME}", "mail", "cn"],
+            attributes=[
+                f"{LDAP_ATTRIBUTE_FOR_USERNAME}",
+                f"{LDAP_ATTRIBUTE_FOR_MAIL}",
+                "cn",
+            ],
        )

        if not search_success:
@@ -226,7 +231,9 @@ async def ldap_auth(request: Request, response: Response, form_data: LdapForm):

        entry = connection_app.entries[0]
        username = str(entry[f"{LDAP_ATTRIBUTE_FOR_USERNAME}"]).lower()
-        mail = str(entry["mail"])
+        mail = str(entry[f"{LDAP_ATTRIBUTE_FOR_MAIL}"])
+        if not mail or mail == "" or mail == "[]":
+            raise HTTPException(400, f"User {form_data.user} does not have mail.")
        cn = str(entry["cn"])
        user_dn = entry.entry_dn

@@ -281,6 +288,10 @@ async def ldap_auth(request: Request, response: Response, form_data: LdapForm):
                    httponly=True,  # Ensures the cookie is not accessible via JavaScript
                )

+                user_permissions = get_permissions(
+                    user.id, request.app.state.config.USER_PERMISSIONS
+                )
+
                return {
                    "token": token,
                    "token_type": "Bearer",
@@ -289,6 +300,7 @@ async def ldap_auth(request: Request, response: Response, form_data: LdapForm):
                    "name": user.name,
                    "role": user.role,
                    "profile_image_url": user.profile_image_url,
+                    "permissions": user_permissions,
                }
            else:
                raise HTTPException(400, detail=ERROR_MESSAGES.INVALID_CRED)
@@ -371,8 +383,8 @@ async def signin(request: Request, response: Response, form_data: SigninForm):
            value=token,
            expires=datetime_expires_at,
            httponly=True,  # Ensures the cookie is not accessible via JavaScript
-            samesite=WEBUI_SESSION_COOKIE_SAME_SITE,
-            secure=WEBUI_SESSION_COOKIE_SECURE,
+            samesite=WEBUI_AUTH_COOKIE_SAME_SITE,
+            secure=WEBUI_AUTH_COOKIE_SECURE,
        )

        user_permissions = get_permissions(
@@ -466,8 +478,8 @@ async def signup(request: Request, response: Response, form_data: SignupForm):
                value=token,
                expires=datetime_expires_at,
                httponly=True,  # Ensures the cookie is not accessible via JavaScript
-                samesite=WEBUI_SESSION_COOKIE_SAME_SITE,
-                secure=WEBUI_SESSION_COOKIE_SECURE,
+                samesite=WEBUI_AUTH_COOKIE_SAME_SITE,
+                secure=WEBUI_AUTH_COOKIE_SECURE,
            )

            if request.app.state.config.WEBHOOK_URL:
@@ -691,6 +703,7 @@ class LdapServerConfig(BaseModel):
    label: str
    host: str
    port: Optional[int] = None
+    attribute_for_mail: str = "mail"
    attribute_for_username: str = "uid"
    app_dn: str
    app_dn_password: str
@@ -707,6 +720,7 @@ async def get_ldap_server(request: Request, user=Depends(get_admin_user)):
        "label": request.app.state.config.LDAP_SERVER_LABEL,
        "host": request.app.state.config.LDAP_SERVER_HOST,
        "port": request.app.state.config.LDAP_SERVER_PORT,
+        "attribute_for_mail": request.app.state.config.LDAP_ATTRIBUTE_FOR_MAIL,
        "attribute_for_username": request.app.state.config.LDAP_ATTRIBUTE_FOR_USERNAME,
        "app_dn": request.app.state.config.LDAP_APP_DN,
        "app_dn_password": request.app.state.config.LDAP_APP_PASSWORD,
@@ -725,6 +739,7 @@ async def update_ldap_server(
    required_fields = [
        "label",
        "host",
+        "attribute_for_mail",
        "attribute_for_username",
        "app_dn",
        "app_dn_password",
@@ -743,6 +758,7 @@ async def update_ldap_server(
    request.app.state.config.LDAP_SERVER_LABEL = form_data.label
    request.app.state.config.LDAP_SERVER_HOST = form_data.host
    request.app.state.config.LDAP_SERVER_PORT = form_data.port
+    request.app.state.config.LDAP_ATTRIBUTE_FOR_MAIL = form_data.attribute_for_mail
    request.app.state.config.LDAP_ATTRIBUTE_FOR_USERNAME = (
        form_data.attribute_for_username
    )
@@ -758,6 +774,7 @@ async def update_ldap_server(
        "label": request.app.state.config.LDAP_SERVER_LABEL,
        "host": request.app.state.config.LDAP_SERVER_HOST,
        "port": request.app.state.config.LDAP_SERVER_PORT,
+        "attribute_for_mail": request.app.state.config.LDAP_ATTRIBUTE_FOR_MAIL,
        "attribute_for_username": request.app.state.config.LDAP_ATTRIBUTE_FOR_USERNAME,
        "app_dn": request.app.state.config.LDAP_APP_DN,
        "app_dn_password": request.app.state.config.LDAP_APP_PASSWORD,
@@ -11,7 +11,12 @@ from open_webui.socket.main import sio, get_user_ids_from_room
 from open_webui.models.users import Users, UserNameResponse

 from open_webui.models.channels import Channels, ChannelModel, ChannelForm
-from open_webui.models.messages import Messages, MessageModel, MessageForm
+from open_webui.models.messages import (
+    Messages,
+    MessageModel,
+    MessageResponse,
+    MessageForm,
+)


 from open_webui.config import ENABLE_ADMIN_CHAT_ACCESS, ENABLE_ADMIN_EXPORT
@@ -49,7 +54,7 @@ async def get_channels(user=Depends(get_verified_user)):
@router.post("/create", response_model=Optional[ChannelModel])
 async def create_new_channel(form_data: ChannelForm, user=Depends(get_admin_user)):
    try:
-        channel = Channels.insert_new_channel(form_data, user.id)
+        channel = Channels.insert_new_channel(None, form_data, user.id)
        return ChannelModel(**channel.model_dump())
    except Exception as e:
        log.exception(e)
@@ -134,11 +139,11 @@ async def delete_channel_by_id(id: str, user=Depends(get_admin_user)):
 ############################


-class MessageUserModel(MessageModel):
+class MessageUserResponse(MessageResponse):
    user: UserNameResponse


-@router.get("/{id}/messages", response_model=list[MessageUserModel])
+@router.get("/{id}/messages", response_model=list[MessageUserResponse])
 async def get_channel_messages(
    id: str, skip: int = 0, limit: int = 50, user=Depends(get_verified_user)
 ):
@@ -164,10 +169,16 @@ async def get_channel_messages(
            user = Users.get_user_by_id(message.user_id)
            users[message.user_id] = user

+        replies = Messages.get_replies_by_message_id(message.id)
+        latest_reply_at = replies[0].created_at if replies else None
+
        messages.append(
-            MessageUserModel(
+            MessageUserResponse(
                **{
                    **message.model_dump(),
+                    "reply_count": len(replies),
+                    "latest_reply_at": latest_reply_at,
+                    "reactions": Messages.get_reactions_by_message_id(message.id),
                    "user": UserNameResponse(**users[message.user_id].model_dump()),
                }
            )
@@ -236,10 +247,17 @@ async def post_new_message(
                "message_id": message.id,
                "data": {
                    "type": "message",
-                    "data": {
-                        **message.model_dump(),
-                        "user": UserNameResponse(**user.model_dump()).model_dump(),
-                    },
+                    "data": MessageUserResponse(
+                        **{
+                            **message.model_dump(),
+                            "reply_count": 0,
+                            "latest_reply_at": None,
+                            "reactions": Messages.get_reactions_by_message_id(
+                                message.id
+                            ),
+                            "user": UserNameResponse(**user.model_dump()),
+                        }
+                    ).model_dump(),
                },
                "user": UserNameResponse(**user.model_dump()).model_dump(),
                "channel": channel.model_dump(),
@@ -251,6 +269,35 @@ async def post_new_message(
                to=f"channel:{channel.id}",
            )

+            if message.parent_id:
+                # If this message is a reply, emit to the parent message as well
+                parent_message = Messages.get_message_by_id(message.parent_id)
+
+                if parent_message:
+                    await sio.emit(
+                        "channel-events",
+                        {
+                            "channel_id": channel.id,
+                            "message_id": parent_message.id,
+                            "data": {
+                                "type": "message:reply",
+                                "data": MessageUserResponse(
+                                    **{
+                                        **parent_message.model_dump(),
+                                        "user": UserNameResponse(
+                                            **Users.get_user_by_id(
+                                                parent_message.user_id
+                                            ).model_dump()
+                                        ),
+                                    }
+                                ).model_dump(),
+                            },
+                            "user": UserNameResponse(**user.model_dump()).model_dump(),
+                            "channel": channel.model_dump(),
+                        },
+                        to=f"channel:{channel.id}",
+                    )
+
            active_user_ids = get_user_ids_from_room(f"channel:{channel.id}")

            background_tasks.add_task(
@@ -269,6 +316,101 @@ async def post_new_message(
        )


+############################
+# GetChannelMessage
+############################
+
+
+@router.get("/{id}/messages/{message_id}", response_model=Optional[MessageUserResponse])
+async def get_channel_message(
+    id: str, message_id: str, user=Depends(get_verified_user)
+):
+    channel = Channels.get_channel_by_id(id)
+    if not channel:
+        raise HTTPException(
+            status_code=status.HTTP_404_NOT_FOUND, detail=ERROR_MESSAGES.NOT_FOUND
+        )
+
+    if user.role != "admin" and not has_access(
+        user.id, type="read", access_control=channel.access_control
+    ):
+        raise HTTPException(
+            status_code=status.HTTP_403_FORBIDDEN, detail=ERROR_MESSAGES.DEFAULT()
+        )
+
+    message = Messages.get_message_by_id(message_id)
+    if not message:
+        raise HTTPException(
+            status_code=status.HTTP_404_NOT_FOUND, detail=ERROR_MESSAGES.NOT_FOUND
+        )
+
+    if message.channel_id != id:
+        raise HTTPException(
+            status_code=status.HTTP_400_BAD_REQUEST, detail=ERROR_MESSAGES.DEFAULT()
+        )
+
+    return MessageUserResponse(
+        **{
+            **message.model_dump(),
+            "user": UserNameResponse(
+                **Users.get_user_by_id(message.user_id).model_dump()
+            ),
+        }
+    )
+
+
+############################
+# GetChannelThreadMessages
+############################
+
+
+@router.get(
+    "/{id}/messages/{message_id}/thread", response_model=list[MessageUserResponse]
+)
+async def get_channel_thread_messages(
+    id: str,
+    message_id: str,
+    skip: int = 0,
+    limit: int = 50,
+    user=Depends(get_verified_user),
+):
+    channel = Channels.get_channel_by_id(id)
+    if not channel:
+        raise HTTPException(
+            status_code=status.HTTP_404_NOT_FOUND, detail=ERROR_MESSAGES.NOT_FOUND
+        )
+
+    if user.role != "admin" and not has_access(
+        user.id, type="read", access_control=channel.access_control
+    ):
+        raise HTTPException(
+            status_code=status.HTTP_403_FORBIDDEN, detail=ERROR_MESSAGES.DEFAULT()
+        )
+
+    message_list = Messages.get_messages_by_parent_id(id, message_id, skip, limit)
+    users = {}
+
+    messages = []
+    for message in message_list:
+        if message.user_id not in users:
+            user = Users.get_user_by_id(message.user_id)
+            users[message.user_id] = user
+
+        messages.append(
+            MessageUserResponse(
+                **{
+                    **message.model_dump(),
+                    "reply_count": 0,
+                    "latest_reply_at": None,
+                    "reactions": Messages.get_reactions_by_message_id(message.id),
+                    "user": UserNameResponse(**users[message.user_id].model_dump()),
+                }
+            )
+        )
+
+    return messages
+
+
 ############################
 # UpdateMessageById
 ############################
@@ -306,6 +448,8 @@ async def update_message_by_id(

    try:
        message = Messages.update_message_by_id(message_id, form_data)
+        message = Messages.get_message_by_id(message_id)
+
        if message:
            await sio.emit(
                "channel-events",
@@ -314,10 +458,14 @@ async def update_message_by_id(
                    "message_id": message.id,
                    "data": {
                        "type": "message:update",
-                        "data": {
-                            **message.model_dump(),
-                            "user": UserNameResponse(**user.model_dump()).model_dump(),
-                        },
+                        "data": MessageUserResponse(
+                            **{
+                                **message.model_dump(),
+                                "user": UserNameResponse(
+                                    **user.model_dump()
+                                ).model_dump(),
+                            }
+                        ).model_dump(),
                    },
                    "user": UserNameResponse(**user.model_dump()).model_dump(),
                    "channel": channel.model_dump(),
@@ -333,6 +481,145 @@ async def update_message_by_id(
        )


+############################
+# AddReactionToMessage
+############################
+
+
+class ReactionForm(BaseModel):
+    name: str
+
+
+@router.post("/{id}/messages/{message_id}/reactions/add", response_model=bool)
+async def add_reaction_to_message(
+    id: str, message_id: str, form_data: ReactionForm, user=Depends(get_verified_user)
+):
+    channel = Channels.get_channel_by_id(id)
+    if not channel:
+        raise HTTPException(
+            status_code=status.HTTP_404_NOT_FOUND, detail=ERROR_MESSAGES.NOT_FOUND
+        )
+
+    if user.role != "admin" and not has_access(
+        user.id, type="read", access_control=channel.access_control
+    ):
+        raise HTTPException(
+            status_code=status.HTTP_403_FORBIDDEN, detail=ERROR_MESSAGES.DEFAULT()
+        )
+
+    message = Messages.get_message_by_id(message_id)
+    if not message:
+        raise HTTPException(
+            status_code=status.HTTP_404_NOT_FOUND, detail=ERROR_MESSAGES.NOT_FOUND
+        )
+
+    if message.channel_id != id:
+        raise HTTPException(
+            status_code=status.HTTP_400_BAD_REQUEST, detail=ERROR_MESSAGES.DEFAULT()
+        )
+
+    try:
+        Messages.add_reaction_to_message(message_id, user.id, form_data.name)
+        message = Messages.get_message_by_id(message_id)
+
+        await sio.emit(
+            "channel-events",
+            {
+                "channel_id": channel.id,
+                "message_id": message.id,
+                "data": {
+                    "type": "message:reaction:add",
+                    "data": {
+                        **message.model_dump(),
+                        "user": UserNameResponse(
+                            **Users.get_user_by_id(message.user_id).model_dump()
+                        ).model_dump(),
+                        "name": form_data.name,
+                    },
+                },
+                "user": UserNameResponse(**user.model_dump()).model_dump(),
+                "channel": channel.model_dump(),
+            },
+            to=f"channel:{channel.id}",
+        )
+
+        return True
+    except Exception as e:
+        log.exception(e)
+        raise HTTPException(
+            status_code=status.HTTP_400_BAD_REQUEST, detail=ERROR_MESSAGES.DEFAULT()
+        )
+
+
+############################
+# RemoveReactionById
+############################
+
+
+@router.post("/{id}/messages/{message_id}/reactions/remove", response_model=bool)
+async def remove_reaction_by_id_and_user_id_and_name(
+    id: str, message_id: str, form_data: ReactionForm, user=Depends(get_verified_user)
+):
+    channel = Channels.get_channel_by_id(id)
+    if not channel:
+        raise HTTPException(
+            status_code=status.HTTP_404_NOT_FOUND, detail=ERROR_MESSAGES.NOT_FOUND
+        )
+
+    if user.role != "admin" and not has_access(
+        user.id, type="read", access_control=channel.access_control
+    ):
+        raise HTTPException(
+            status_code=status.HTTP_403_FORBIDDEN, detail=ERROR_MESSAGES.DEFAULT()
+        )
+
+    message = Messages.get_message_by_id(message_id)
+    if not message:
+        raise HTTPException(
+            status_code=status.HTTP_404_NOT_FOUND, detail=ERROR_MESSAGES.NOT_FOUND
+        )
+
+    if message.channel_id != id:
+        raise HTTPException(
+            status_code=status.HTTP_400_BAD_REQUEST, detail=ERROR_MESSAGES.DEFAULT()
+        )
+
+    try:
+        Messages.remove_reaction_by_id_and_user_id_and_name(
+            message_id, user.id, form_data.name
+        )
+
+        message = Messages.get_message_by_id(message_id)
+
+        await sio.emit(
+            "channel-events",
+            {
+                "channel_id": channel.id,
+                "message_id": message.id,
+                "data": {
+                    "type": "message:reaction:remove",
+                    "data": {
+                        **message.model_dump(),
+                        "user": UserNameResponse(
+                            **Users.get_user_by_id(message.user_id).model_dump()
+                        ).model_dump(),
+                        "name": form_data.name,
+                    },
+                },
+                "user": UserNameResponse(**user.model_dump()).model_dump(),
+                "channel": channel.model_dump(),
+            },
+            to=f"channel:{channel.id}",
+        )
+
+        return True
+    except Exception as e:
+        log.exception(e)
+        raise HTTPException(
+            status_code=status.HTTP_400_BAD_REQUEST, detail=ERROR_MESSAGES.DEFAULT()
+        )
+
+
 ############################
 # DeleteMessageById
 ############################
@@ -386,6 +673,35 @@ async def delete_message_by_id(
            to=f"channel:{channel.id}",
        )

+        if message.parent_id:
+            # If this message is a reply, emit to the parent message as well
+            parent_message = Messages.get_message_by_id(message.parent_id)
+
+            if parent_message:
+                await sio.emit(
+                    "channel-events",
+                    {
+                        "channel_id": channel.id,
+                        "message_id": parent_message.id,
+                        "data": {
+                            "type": "message:reply",
+                            "data": MessageUserResponse(
+                                **{
+                                    **parent_message.model_dump(),
+                                    "user": UserNameResponse(
+                                        **Users.get_user_by_id(
+                                            parent_message.user_id
+                                        ).model_dump()
+                                    ),
+                                }
+                            ).model_dump(),
+                        },
+                        "user": UserNameResponse(**user.model_dump()).model_dump(),
+                        "channel": channel.model_dump(),
+                    },
+                    to=f"channel:{channel.id}",
+                )
+
        return True
    except Exception as e:
        log.exception(e)
@@ -444,9 +444,39 @@ async def pin_chat_by_id(id: str, user=Depends(get_verified_user)):
 ############################


+class CloneForm(BaseModel):
+    title: Optional[str] = None
+
+
@router.post("/{id}/clone", response_model=Optional[ChatResponse])
-async def clone_chat_by_id(id: str, user=Depends(get_verified_user)):
+async def clone_chat_by_id(
+    form_data: CloneForm, id: str, user=Depends(get_verified_user)
+):
    chat = Chats.get_chat_by_id_and_user_id(id, user.id)
+    if chat:
+        updated_chat = {
+            **chat.chat,
+            "originalChatId": chat.id,
+            "branchPointMessageId": chat.chat["history"]["currentId"],
+            "title": form_data.title if form_data.title else f"Clone of {chat.title}",
+        }
+
+        chat = Chats.insert_new_chat(user.id, ChatForm(**{"chat": updated_chat}))
+        return ChatResponse(**chat.model_dump())
+    else:
+        raise HTTPException(
+            status_code=status.HTTP_401_UNAUTHORIZED, detail=ERROR_MESSAGES.DEFAULT()
+        )
+
+
+############################
+# CloneSharedChatById
+############################
+
+
+@router.post("/{id}/clone/shared", response_model=Optional[ChatResponse])
+async def clone_shared_chat_by_id(id: str, user=Depends(get_verified_user)):
+    chat = Chats.get_chat_by_share_id(id)
    if chat:
        updated_chat = {
            **chat.chat,
@@ -71,7 +71,7 @@ def upload_file(
        )

        try:
-            process_file(request, ProcessFileForm(file_id=id))
+            process_file(request, ProcessFileForm(file_id=id), user=user)
            file_item = Files.get_file_by_id(id=id)
        except Exception as e:
            log.exception(e)
@@ -193,7 +193,9 @@ async def update_file_data_content_by_id(
    if file and (file.user_id == user.id or user.role == "admin"):
        try:
            process_file(
-                request, ProcessFileForm(file_id=id, content=form_data.content)
+                request,
+                ProcessFileForm(file_id=id, content=form_data.content),
+                user=user,
            )
            file = Files.get_file_by_id(id=id)
        except Exception as e:
@@ -345,10 +347,12 @@ async def get_file_content_by_id(id: str, user=Depends(get_verified_user)):
 async def delete_file_by_id(id: str, user=Depends(get_verified_user)):
    file = Files.get_file_by_id(id)
    if file and (file.user_id == user.id or user.role == "admin"):
+        # We should add Chroma cleanup here
+
        result = Files.delete_file_by_id(id)
        if result:
            try:
-                Storage.delete_file(file.filename)
+                Storage.delete_file(file.path)
            except Exception as e:
                log.exception(e)
                log.error(f"Error deleting files")
@@ -2,6 +2,8 @@ import os
 from pathlib import Path
 from typing import Optional

+
+from open_webui.models.users import Users
 from open_webui.models.groups import (
    Groups,
    GroupForm,
@@ -80,6 +82,9 @@ async def update_group_by_id(
    id: str, form_data: GroupUpdateForm, user=Depends(get_admin_user)
 ):
    try:
+        if form_data.user_ids:
+            form_data.user_ids = Users.get_valid_user_ids(form_data.user_ids)
+
        group = Groups.update_group_by_id(id, form_data)
        if group:
            return group
@@ -43,6 +43,7 @@ async def get_config(request: Request, user=Depends(get_admin_user)):
    return {
        "enabled": request.app.state.config.ENABLE_IMAGE_GENERATION,
        "engine": request.app.state.config.IMAGE_GENERATION_ENGINE,
+        "prompt_generation": request.app.state.config.ENABLE_IMAGE_PROMPT_GENERATION,
        "openai": {
            "OPENAI_API_BASE_URL": request.app.state.config.IMAGES_OPENAI_API_BASE_URL,
            "OPENAI_API_KEY": request.app.state.config.IMAGES_OPENAI_API_KEY,
@@ -86,6 +87,7 @@ class ComfyUIConfigForm(BaseModel):
 class ConfigForm(BaseModel):
    enabled: bool
    engine: str
+    prompt_generation: bool
    openai: OpenAIConfigForm
    automatic1111: Automatic1111ConfigForm
    comfyui: ComfyUIConfigForm
@@ -98,6 +100,10 @@ async def update_config(
    request.app.state.config.IMAGE_GENERATION_ENGINE = form_data.engine
    request.app.state.config.ENABLE_IMAGE_GENERATION = form_data.enabled

+    request.app.state.config.ENABLE_IMAGE_PROMPT_GENERATION = (
+        form_data.prompt_generation
+    )
+
    request.app.state.config.IMAGES_OPENAI_API_BASE_URL = (
        form_data.openai.OPENAI_API_BASE_URL
    )
@@ -137,6 +143,7 @@ async def update_config(
    return {
        "enabled": request.app.state.config.ENABLE_IMAGE_GENERATION,
        "engine": request.app.state.config.IMAGE_GENERATION_ENGINE,
+        "prompt_generation": request.app.state.config.ENABLE_IMAGE_PROMPT_GENERATION,
        "openai": {
            "OPENAI_API_BASE_URL": request.app.state.config.IMAGES_OPENAI_API_BASE_URL,
            "OPENAI_API_KEY": request.app.state.config.IMAGES_OPENAI_API_KEY,
@@ -408,10 +415,14 @@ def save_b64_image(b64_str):
        return None


-def save_url_image(url):
+def save_url_image(url, headers=None):
    image_id = str(uuid.uuid4())
    try:
-        r = requests.get(url)
+        if headers:
+            r = requests.get(url, headers=headers)
+        else:
+            r = requests.get(url)
+
        r.raise_for_status()
        if r.headers["content-type"].split("/")[0] == "image":
            mime_type = r.headers["content-type"]
@@ -535,7 +546,13 @@ async def image_generations(
            images = []

            for image in res["data"]:
-                image_filename = save_url_image(image["url"])
+                headers = None
+                if request.app.state.config.COMFYUI_API_KEY:
+                    headers = {
+                        "Authorization": f"Bearer {request.app.state.config.COMFYUI_API_KEY}"
+                    }
+
+                image_filename = save_url_image(image["url"], headers)
                images.append({"url": f"/cache/image/generations/{image_filename}"})
                file_body_path = IMAGE_CACHE_DIR.joinpath(f"{image_filename}.json")

@@ -17,7 +17,7 @@ from open_webui.routers.retrieval import (
    process_files_batch,
    BatchProcessFilesForm,
 )
-
+from open_webui.storage.provider import Storage

 from open_webui.constants import ERROR_MESSAGES
 from open_webui.utils.auth import get_verified_user
@@ -25,6 +25,7 @@ from open_webui.utils.access_control import has_access, has_permission


 from open_webui.env import SRC_LOG_LEVELS
+from open_webui.models.models import Models, ModelForm


 log = logging.getLogger(__name__)
@@ -212,8 +213,12 @@ async def update_knowledge_by_id(
            status_code=status.HTTP_400_BAD_REQUEST,
            detail=ERROR_MESSAGES.NOT_FOUND,
        )
-
-    if knowledge.user_id != user.id and user.role != "admin":
+    # Is the user the original creator, in a group with write access, or an admin
+    if (
+        knowledge.user_id != user.id
+        and not has_access(user.id, "write", knowledge.access_control)
+        and user.role != "admin"
+    ):
        raise HTTPException(
            status_code=status.HTTP_400_BAD_REQUEST,
            detail=ERROR_MESSAGES.ACCESS_PROHIBITED,
@@ -259,7 +264,11 @@ def add_file_to_knowledge_by_id(
            detail=ERROR_MESSAGES.NOT_FOUND,
        )

-    if knowledge.user_id != user.id and user.role != "admin":
+    if (
+        knowledge.user_id != user.id
+        and not has_access(user.id, "write", knowledge.access_control)
+        and user.role != "admin"
+    ):
        raise HTTPException(
            status_code=status.HTTP_400_BAD_REQUEST,
            detail=ERROR_MESSAGES.ACCESS_PROHIBITED,
@@ -280,7 +289,9 @@ def add_file_to_knowledge_by_id(
    # Add content to the vector database
    try:
        process_file(
-            request, ProcessFileForm(file_id=form_data.file_id, collection_name=id)
+            request,
+            ProcessFileForm(file_id=form_data.file_id, collection_name=id),
+            user=user,
        )
    except Exception as e:
        log.debug(e)
@@ -337,7 +348,12 @@ def update_file_from_knowledge_by_id(
            detail=ERROR_MESSAGES.NOT_FOUND,
        )

-    if knowledge.user_id != user.id and user.role != "admin":
+    if (
+        knowledge.user_id != user.id
+        and not has_access(user.id, "write", knowledge.access_control)
+        and user.role != "admin"
+    ):
+
        raise HTTPException(
            status_code=status.HTTP_400_BAD_REQUEST,
            detail=ERROR_MESSAGES.ACCESS_PROHIBITED,
@@ -358,7 +374,9 @@ def update_file_from_knowledge_by_id(
    # Add content to the vector database
    try:
        process_file(
-            request, ProcessFileForm(file_id=form_data.file_id, collection_name=id)
+            request,
+            ProcessFileForm(file_id=form_data.file_id, collection_name=id),
+            user=user,
        )
    except Exception as e:
        raise HTTPException(
@@ -401,7 +419,11 @@ def remove_file_from_knowledge_by_id(
            detail=ERROR_MESSAGES.NOT_FOUND,
        )

-    if knowledge.user_id != user.id and user.role != "admin":
+    if (
+        knowledge.user_id != user.id
+        and not has_access(user.id, "write", knowledge.access_control)
+        and user.role != "admin"
+    ):
        raise HTTPException(
            status_code=status.HTTP_400_BAD_REQUEST,
            detail=ERROR_MESSAGES.ACCESS_PROHIBITED,
@@ -419,11 +441,12 @@ def remove_file_from_knowledge_by_id(
        collection_name=knowledge.id, filter={"file_id": form_data.file_id}
    )

-    result = VECTOR_DB_CLIENT.query(
-        collection_name=knowledge.id,
-        filter={"file_id": form_data.file_id},
-    )
+    # Remove the file's collection from vector database
+    file_collection = f"file-{form_data.file_id}"
+    if VECTOR_DB_CLIENT.has_collection(collection_name=file_collection):
+        VECTOR_DB_CLIENT.delete_collection(collection_name=file_collection)

+    # Delete file from database
    Files.delete_file_by_id(form_data.file_id)

    if knowledge:
@@ -474,12 +497,46 @@ async def delete_knowledge_by_id(id: str, user=Depends(get_verified_user)):
            detail=ERROR_MESSAGES.NOT_FOUND,
        )

-    if knowledge.user_id != user.id and user.role != "admin":
+    if (
+        knowledge.user_id != user.id
+        and not has_access(user.id, "write", knowledge.access_control)
+        and user.role != "admin"
+    ):
        raise HTTPException(
            status_code=status.HTTP_400_BAD_REQUEST,
            detail=ERROR_MESSAGES.ACCESS_PROHIBITED,
        )

+    log.info(f"Deleting knowledge base: {id} (name: {knowledge.name})")
+
+    # Get all models
+    models = Models.get_all_models()
+    log.info(f"Found {len(models)} models to check for knowledge base {id}")
+
+    # Update models that reference this knowledge base
+    for model in models:
+        if model.meta and hasattr(model.meta, "knowledge"):
+            knowledge_list = model.meta.knowledge or []
+            # Filter out the deleted knowledge base
+            updated_knowledge = [k for k in knowledge_list if k.get("id") != id]
+
+            # If the knowledge list changed, update the model
+            if len(updated_knowledge) != len(knowledge_list):
+                log.info(f"Updating model {model.id} to remove knowledge base {id}")
+                model.meta.knowledge = updated_knowledge
+                # Create a ModelForm for the update
+                model_form = ModelForm(
+                    id=model.id,
+                    name=model.name,
+                    base_model_id=model.base_model_id,
+                    meta=model.meta,
+                    params=model.params,
+                    access_control=model.access_control,
+                    is_active=model.is_active,
+                )
+                Models.update_model_by_id(model.id, model_form)
+
+    # Clean up vector DB
    try:
        VECTOR_DB_CLIENT.delete_collection(collection_name=id)
    except Exception as e:
@@ -503,7 +560,11 @@ async def reset_knowledge_by_id(id: str, user=Depends(get_verified_user)):
            detail=ERROR_MESSAGES.NOT_FOUND,
        )

-    if knowledge.user_id != user.id and user.role != "admin":
+    if (
+        knowledge.user_id != user.id
+        and not has_access(user.id, "write", knowledge.access_control)
+        and user.role != "admin"
+    ):
        raise HTTPException(
            status_code=status.HTTP_400_BAD_REQUEST,
            detail=ERROR_MESSAGES.ACCESS_PROHIBITED,
@@ -527,6 +588,7 @@ async def reset_knowledge_by_id(id: str, user=Depends(get_verified_user)):

@router.post("/{id}/files/batch/add", response_model=Optional[KnowledgeFilesResponse])
 def add_files_to_knowledge_batch(
+    request: Request,
    id: str,
    form_data: list[KnowledgeFileIdForm],
    user=Depends(get_verified_user),
@@ -541,7 +603,11 @@ def add_files_to_knowledge_batch(
            detail=ERROR_MESSAGES.NOT_FOUND,
        )

-    if knowledge.user_id != user.id and user.role != "admin":
+    if (
+        knowledge.user_id != user.id
+        and not has_access(user.id, "write", knowledge.access_control)
+        and user.role != "admin"
+    ):
        raise HTTPException(
            status_code=status.HTTP_400_BAD_REQUEST,
            detail=ERROR_MESSAGES.ACCESS_PROHIBITED,
@@ -562,7 +628,9 @@ def add_files_to_knowledge_batch(
    # Process files
    try:
        result = process_files_batch(
-            BatchProcessFilesForm(files=files, collection_name=id)
+            request=request,
+            form_data=BatchProcessFilesForm(files=files, collection_name=id),
+            user=user,
        )
    except Exception as e:
        log.error(
@@ -57,7 +57,7 @@ async def add_memory(
            {
                "id": memory.id,
                "text": memory.content,
-                "vector": request.app.state.EMBEDDING_FUNCTION(memory.content),
+                "vector": request.app.state.EMBEDDING_FUNCTION(memory.content, user),
                "metadata": {"created_at": memory.created_at},
            }
        ],
@@ -82,7 +82,7 @@ async def query_memory(
 ):
    results = VECTOR_DB_CLIENT.search(
        collection_name=f"user-memory-{user.id}",
-        vectors=[request.app.state.EMBEDDING_FUNCTION(form_data.content)],
+        vectors=[request.app.state.EMBEDDING_FUNCTION(form_data.content, user)],
        limit=form_data.k,
    )

@@ -105,7 +105,7 @@ async def reset_memory_from_vector_db(
            {
                "id": memory.id,
                "text": memory.content,
-                "vector": request.app.state.EMBEDDING_FUNCTION(memory.content),
+                "vector": request.app.state.EMBEDDING_FUNCTION(memory.content, user),
                "metadata": {
                    "created_at": memory.created_at,
                    "updated_at": memory.updated_at,
@@ -160,7 +160,9 @@ async def update_memory_by_id(
                {
                    "id": memory.id,
                    "text": memory.content,
-                    "vector": request.app.state.EMBEDDING_FUNCTION(memory.content),
+                    "vector": request.app.state.EMBEDDING_FUNCTION(
+                        memory.content, user
+                    ),
                    "metadata": {
                        "created_at": memory.created_at,
                        "updated_at": memory.updated_at,
@@ -155,6 +155,16 @@ async def update_model_by_id(
            detail=ERROR_MESSAGES.NOT_FOUND,
        )

+    if (
+        model.user_id != user.id
+        and not has_access(user.id, "write", model.access_control)
+        and user.role != "admin"
+    ):
+        raise HTTPException(
+            status_code=status.HTTP_400_BAD_REQUEST,
+            detail=ERROR_MESSAGES.ACCESS_PROHIBITED,
+        )
+
    model = Models.update_model_by_id(id, form_data)
    return model

@@ -173,7 +183,11 @@ async def delete_model_by_id(id: str, user=Depends(get_verified_user)):
            detail=ERROR_MESSAGES.NOT_FOUND,
        )

-    if model.user_id != user.id and user.role != "admin":
+    if (
+        user.role != "admin"
+        and model.user_id != user.id
+        and not has_access(user.id, "write", model.access_control)
+    ):
        raise HTTPException(
            status_code=status.HTTP_401_UNAUTHORIZED,
            detail=ERROR_MESSAGES.UNAUTHORIZED,
@@ -152,10 +152,12 @@ async def send_post_request(
        )


-def get_api_key(url, configs):
+def get_api_key(idx, url, configs):
    parsed_url = urlparse(url)
    base_url = f"{parsed_url.scheme}://{parsed_url.netloc}"
-    return configs.get(base_url, {}).get("key", None)
+    return configs.get(str(idx), configs.get(base_url, {})).get(
+        "key", None
+    )  # Legacy support


 ##########################################
@@ -238,11 +240,13 @@ async def update_config(
    request.app.state.config.OLLAMA_BASE_URLS = form_data.OLLAMA_BASE_URLS
    request.app.state.config.OLLAMA_API_CONFIGS = form_data.OLLAMA_API_CONFIGS

-    # Remove any extra configs
-    config_urls = request.app.state.config.OLLAMA_API_CONFIGS.keys()
-    for url in list(request.app.state.config.OLLAMA_BASE_URLS):
-        if url not in config_urls:
-            request.app.state.config.OLLAMA_API_CONFIGS.pop(url, None)
+    # Remove the API configs that are not in the API URLS
+    keys = list(map(str, range(len(request.app.state.config.OLLAMA_BASE_URLS))))
+    request.app.state.config.OLLAMA_API_CONFIGS = {
+        key: value
+        for key, value in request.app.state.config.OLLAMA_API_CONFIGS.items()
+        if key in keys
+    }

    return {
        "ENABLE_OLLAMA_API": request.app.state.config.ENABLE_OLLAMA_API,
@@ -256,12 +260,19 @@ async def get_all_models(request: Request):
    log.info("get_all_models()")
    if request.app.state.config.ENABLE_OLLAMA_API:
        request_tasks = []
-
        for idx, url in enumerate(request.app.state.config.OLLAMA_BASE_URLS):
-            if url not in request.app.state.config.OLLAMA_API_CONFIGS:
+            if (str(idx) not in request.app.state.config.OLLAMA_API_CONFIGS) and (
+                url not in request.app.state.config.OLLAMA_API_CONFIGS  # Legacy support
+            ):
                request_tasks.append(send_get_request(f"{url}/api/tags"))
            else:
-                api_config = request.app.state.config.OLLAMA_API_CONFIGS.get(url, {})
+                api_config = request.app.state.config.OLLAMA_API_CONFIGS.get(
+                    str(idx),
+                    request.app.state.config.OLLAMA_API_CONFIGS.get(
+                        url, {}
+                    ),  # Legacy support
+                )
+
                enable = api_config.get("enable", True)
                key = api_config.get("key", None)

@@ -275,7 +286,12 @@ async def get_all_models(request: Request):
        for idx, response in enumerate(responses):
            if response:
                url = request.app.state.config.OLLAMA_BASE_URLS[idx]
-                api_config = request.app.state.config.OLLAMA_API_CONFIGS.get(url, {})
+                api_config = request.app.state.config.OLLAMA_API_CONFIGS.get(
+                    str(idx),
+                    request.app.state.config.OLLAMA_API_CONFIGS.get(
+                        url, {}
+                    ),  # Legacy support
+                )

                prefix_id = api_config.get("prefix_id", None)
                model_ids = api_config.get("model_ids", [])
@@ -349,7 +365,7 @@ async def get_ollama_tags(
        models = await get_all_models(request)
    else:
        url = request.app.state.config.OLLAMA_BASE_URLS[url_idx]
-        key = get_api_key(url, request.app.state.config.OLLAMA_API_CONFIGS)
+        key = get_api_key(url_idx, url, request.app.state.config.OLLAMA_API_CONFIGS)

        r = None
        try:
@@ -379,7 +395,7 @@ async def get_ollama_tags(
            )

    if user.role == "user" and not BYPASS_MODEL_ACCESS_CONTROL:
-        models["models"] = get_filtered_models(models, user)
+        models["models"] = await get_filtered_models(models, user)

    return models

@@ -393,11 +409,14 @@ async def get_ollama_versions(request: Request, url_idx: Optional[int] = None):
            request_tasks = [
                send_get_request(
                    f"{url}/api/version",
-                    request.app.state.config.OLLAMA_API_CONFIGS.get(url, {}).get(
-                        "key", None
-                    ),
+                    request.app.state.config.OLLAMA_API_CONFIGS.get(
+                        str(idx),
+                        request.app.state.config.OLLAMA_API_CONFIGS.get(
+                            url, {}
+                        ),  # Legacy support
+                    ).get("key", None),
                )
-                for url in request.app.state.config.OLLAMA_BASE_URLS
+                for idx, url in enumerate(request.app.state.config.OLLAMA_BASE_URLS)
            ]
            responses = await asyncio.gather(*request_tasks)
            responses = list(filter(lambda x: x is not None, responses))
@@ -454,11 +473,14 @@ async def get_ollama_loaded_models(request: Request, user=Depends(get_verified_u
        request_tasks = [
            send_get_request(
                f"{url}/api/ps",
-                request.app.state.config.OLLAMA_API_CONFIGS.get(url, {}).get(
-                    "key", None
-                ),
+                request.app.state.config.OLLAMA_API_CONFIGS.get(
+                    str(idx),
+                    request.app.state.config.OLLAMA_API_CONFIGS.get(
+                        url, {}
+                    ),  # Legacy support
+                ).get("key", None),
            )
-            for url in request.app.state.config.OLLAMA_BASE_URLS
+            for idx, url in enumerate(request.app.state.config.OLLAMA_BASE_URLS)
        ]
        responses = await asyncio.gather(*request_tasks)

@@ -488,7 +510,7 @@ async def pull_model(
    return await send_post_request(
        url=f"{url}/api/pull",
        payload=json.dumps(payload),
-        key=get_api_key(url, request.app.state.config.OLLAMA_API_CONFIGS),
+        key=get_api_key(url_idx, url, request.app.state.config.OLLAMA_API_CONFIGS),
    )


@@ -524,16 +546,17 @@ async def push_model(
    return await send_post_request(
        url=f"{url}/api/push",
        payload=form_data.model_dump_json(exclude_none=True).encode(),
-        key=get_api_key(url, request.app.state.config.OLLAMA_API_CONFIGS),
+        key=get_api_key(url_idx, url, request.app.state.config.OLLAMA_API_CONFIGS),
    )


 class CreateModelForm(BaseModel):
-    name: str
-    modelfile: Optional[str] = None
+    model: Optional[str] = None
    stream: Optional[bool] = None
    path: Optional[str] = None

+    model_config = ConfigDict(extra="allow")
+

@router.post("/api/create")
@router.post("/api/create/{url_idx}")
@@ -549,7 +572,7 @@ async def create_model(
    return await send_post_request(
        url=f"{url}/api/create",
        payload=form_data.model_dump_json(exclude_none=True).encode(),
-        key=get_api_key(url, request.app.state.config.OLLAMA_API_CONFIGS),
+        key=get_api_key(url_idx, url, request.app.state.config.OLLAMA_API_CONFIGS),
    )


@@ -579,7 +602,7 @@ async def copy_model(
            )

    url = request.app.state.config.OLLAMA_BASE_URLS[url_idx]
-    key = get_api_key(url, request.app.state.config.OLLAMA_API_CONFIGS)
+    key = get_api_key(url_idx, url, request.app.state.config.OLLAMA_API_CONFIGS)

    try:
        r = requests.request(
@@ -634,7 +657,7 @@ async def delete_model(
            )

    url = request.app.state.config.OLLAMA_BASE_URLS[url_idx]
-    key = get_api_key(url, request.app.state.config.OLLAMA_API_CONFIGS)
+    key = get_api_key(url_idx, url, request.app.state.config.OLLAMA_API_CONFIGS)

    try:
        r = requests.request(
@@ -684,7 +707,7 @@ async def show_model_info(
    url_idx = random.choice(models[form_data.name]["urls"])

    url = request.app.state.config.OLLAMA_BASE_URLS[url_idx]
-    key = get_api_key(url, request.app.state.config.OLLAMA_API_CONFIGS)
+    key = get_api_key(url_idx, url, request.app.state.config.OLLAMA_API_CONFIGS)

    try:
        r = requests.request(
@@ -753,7 +776,7 @@ async def embed(
            )

    url = request.app.state.config.OLLAMA_BASE_URLS[url_idx]
-    key = get_api_key(url, request.app.state.config.OLLAMA_API_CONFIGS)
+    key = get_api_key(url_idx, url, request.app.state.config.OLLAMA_API_CONFIGS)

    try:
        r = requests.request(
@@ -822,7 +845,7 @@ async def embeddings(
            )

    url = request.app.state.config.OLLAMA_BASE_URLS[url_idx]
-    key = get_api_key(url, request.app.state.config.OLLAMA_API_CONFIGS)
+    key = get_api_key(url_idx, url, request.app.state.config.OLLAMA_API_CONFIGS)

    try:
        r = requests.request(
@@ -897,7 +920,10 @@ async def generate_completion(
            )

    url = request.app.state.config.OLLAMA_BASE_URLS[url_idx]
-    api_config = request.app.state.config.OLLAMA_API_CONFIGS.get(url, {})
+    api_config = request.app.state.config.OLLAMA_API_CONFIGS.get(
+        str(url_idx),
+        request.app.state.config.OLLAMA_API_CONFIGS.get(url, {}),  # Legacy support
+    )

    prefix_id = api_config.get("prefix_id", None)
    if prefix_id:
@@ -906,13 +932,14 @@ async def generate_completion(
    return await send_post_request(
        url=f"{url}/api/generate",
        payload=form_data.model_dump_json(exclude_none=True).encode(),
-        key=get_api_key(url, request.app.state.config.OLLAMA_API_CONFIGS),
+        key=get_api_key(url_idx, url, request.app.state.config.OLLAMA_API_CONFIGS),
    )


 class ChatMessage(BaseModel):
    role: str
    content: str
+    tool_calls: Optional[list[dict]] = None
    images: Optional[list[str]] = None


@@ -924,6 +951,7 @@ class GenerateChatCompletionForm(BaseModel):
    template: Optional[str] = None
    stream: Optional[bool] = True
    keep_alive: Optional[Union[int, str]] = None
+    tools: Optional[list[dict]] = None


 async def get_ollama_url(request: Request, model: str, url_idx: Optional[int] = None):
@@ -936,7 +964,7 @@ async def get_ollama_url(request: Request, model: str, url_idx: Optional[int] =
            )
        url_idx = random.choice(models[model].get("urls", []))
    url = request.app.state.config.OLLAMA_BASE_URLS[url_idx]
-    return url
+    return url, url_idx


@router.post("/api/chat")
@@ -951,6 +979,7 @@ async def generate_chat_completion(
    if BYPASS_MODEL_ACCESS_CONTROL:
        bypass_filter = True

+    metadata = form_data.pop("metadata", None)
    try:
        form_data = GenerateChatCompletionForm(**form_data)
    except Exception as e:
@@ -961,8 +990,6 @@ async def generate_chat_completion(
        )

    payload = {**form_data.model_dump(exclude_none=True)}
-    if "metadata" in payload:
-        del payload["metadata"]

    model_id = payload["model"]
    model_info = Models.get_model_by_id(model_id)
@@ -980,7 +1007,7 @@ async def generate_chat_completion(
            payload["options"] = apply_model_params_to_body_ollama(
                params, payload["options"]
            )
-            payload = apply_model_system_prompt_to_body(params, payload, user)
+            payload = apply_model_system_prompt_to_body(params, payload, metadata)

        # Check if user has access to the model
        if not bypass_filter and user.role == "user":
@@ -1004,8 +1031,11 @@ async def generate_chat_completion(
    if ":" not in payload["model"]:
        payload["model"] = f"{payload['model']}:latest"

-    url = await get_ollama_url(request, payload["model"], url_idx)
-    api_config = request.app.state.config.OLLAMA_API_CONFIGS.get(url, {})
+    url, url_idx = await get_ollama_url(request, payload["model"], url_idx)
+    api_config = request.app.state.config.OLLAMA_API_CONFIGS.get(
+        str(url_idx),
+        request.app.state.config.OLLAMA_API_CONFIGS.get(url, {}),  # Legacy support
+    )

    prefix_id = api_config.get("prefix_id", None)
    if prefix_id:
@@ -1015,7 +1045,7 @@ async def generate_chat_completion(
        url=f"{url}/api/chat",
        payload=json.dumps(payload),
        stream=form_data.stream,
-        key=get_api_key(url, request.app.state.config.OLLAMA_API_CONFIGS),
+        key=get_api_key(url_idx, url, request.app.state.config.OLLAMA_API_CONFIGS),
        content_type="application/x-ndjson",
    )

@@ -1103,8 +1133,11 @@ async def generate_openai_completion(
    if ":" not in payload["model"]:
        payload["model"] = f"{payload['model']}:latest"

-    url = await get_ollama_url(request, payload["model"], url_idx)
-    api_config = request.app.state.config.OLLAMA_API_CONFIGS.get(url, {})
+    url, url_idx = await get_ollama_url(request, payload["model"], url_idx)
+    api_config = request.app.state.config.OLLAMA_API_CONFIGS.get(
+        str(url_idx),
+        request.app.state.config.OLLAMA_API_CONFIGS.get(url, {}),  # Legacy support
+    )

    prefix_id = api_config.get("prefix_id", None)

@@ -1115,7 +1148,7 @@ async def generate_openai_completion(
        url=f"{url}/v1/completions",
        payload=json.dumps(payload),
        stream=payload.get("stream", False),
-        key=get_api_key(url, request.app.state.config.OLLAMA_API_CONFIGS),
+        key=get_api_key(url_idx, url, request.app.state.config.OLLAMA_API_CONFIGS),
    )


@@ -1177,8 +1210,11 @@ async def generate_openai_chat_completion(
    if ":" not in payload["model"]:
        payload["model"] = f"{payload['model']}:latest"

-    url = await get_ollama_url(request, payload["model"], url_idx)
-    api_config = request.app.state.config.OLLAMA_API_CONFIGS.get(url, {})
+    url, url_idx = await get_ollama_url(request, payload["model"], url_idx)
+    api_config = request.app.state.config.OLLAMA_API_CONFIGS.get(
+        str(url_idx),
+        request.app.state.config.OLLAMA_API_CONFIGS.get(url, {}),  # Legacy support
+    )

    prefix_id = api_config.get("prefix_id", None)
    if prefix_id:
@@ -1188,7 +1224,7 @@ async def generate_openai_chat_completion(
        url=f"{url}/v1/chat/completions",
        payload=json.dumps(payload),
        stream=payload.get("stream", False),
-        key=get_api_key(url, request.app.state.config.OLLAMA_API_CONFIGS),
+        key=get_api_key(url_idx, url, request.app.state.config.OLLAMA_API_CONFIGS),
    )


@@ -145,11 +145,13 @@ async def update_config(

    request.app.state.config.OPENAI_API_CONFIGS = form_data.OPENAI_API_CONFIGS

-    # Remove any extra configs
-    config_urls = request.app.state.config.OPENAI_API_CONFIGS.keys()
-    for idx, url in enumerate(request.app.state.config.OPENAI_API_BASE_URLS):
-        if url not in config_urls:
-            request.app.state.config.OPENAI_API_CONFIGS.pop(url, None)
+    # Remove the API configs that are not in the API URLS
+    keys = list(map(str, range(len(request.app.state.config.OPENAI_API_BASE_URLS))))
+    request.app.state.config.OPENAI_API_CONFIGS = {
+        key: value
+        for key, value in request.app.state.config.OPENAI_API_CONFIGS.items()
+        if key in keys
+    }

    return {
        "ENABLE_OPENAI_API": request.app.state.config.ENABLE_OPENAI_API,
@@ -264,14 +266,21 @@ async def get_all_models_responses(request: Request) -> list:

    request_tasks = []
    for idx, url in enumerate(request.app.state.config.OPENAI_API_BASE_URLS):
-        if url not in request.app.state.config.OPENAI_API_CONFIGS:
+        if (str(idx) not in request.app.state.config.OPENAI_API_CONFIGS) and (
+            url not in request.app.state.config.OPENAI_API_CONFIGS  # Legacy support
+        ):
            request_tasks.append(
                send_get_request(
                    f"{url}/models", request.app.state.config.OPENAI_API_KEYS[idx]
                )
            )
        else:
-            api_config = request.app.state.config.OPENAI_API_CONFIGS.get(url, {})
+            api_config = request.app.state.config.OPENAI_API_CONFIGS.get(
+                str(idx),
+                request.app.state.config.OPENAI_API_CONFIGS.get(
+                    url, {}
+                ),  # Legacy support
+            )

            enable = api_config.get("enable", True)
            model_ids = api_config.get("model_ids", [])
@@ -310,7 +319,12 @@ async def get_all_models_responses(request: Request) -> list:
    for idx, response in enumerate(responses):
        if response:
            url = request.app.state.config.OPENAI_API_BASE_URLS[idx]
-            api_config = request.app.state.config.OPENAI_API_CONFIGS.get(url, {})
+            api_config = request.app.state.config.OPENAI_API_CONFIGS.get(
+                str(idx),
+                request.app.state.config.OPENAI_API_CONFIGS.get(
+                    url, {}
+                ),  # Legacy support
+            )

            prefix_id = api_config.get("prefix_id", None)

@@ -475,7 +489,7 @@ async def get_models(
                raise HTTPException(status_code=500, detail=error_detail)

    if user.role == "user" and not BYPASS_MODEL_ACCESS_CONTROL:
-        models["data"] = get_filtered_models(models, user)
+        models["data"] = await get_filtered_models(models, user)

    return models

@@ -537,9 +551,9 @@ async def generate_chat_completion(
        bypass_filter = True

    idx = 0
+
    payload = {**form_data}
-    if "metadata" in payload:
-        del payload["metadata"]
+    metadata = payload.pop("metadata", None)

    model_id = form_data.get("model")
    model_info = Models.get_model_by_id(model_id)
@@ -552,7 +566,7 @@ async def generate_chat_completion(

        params = model_info.params.model_dump()
        payload = apply_model_params_to_body_openai(params, payload)
-        payload = apply_model_system_prompt_to_body(params, payload, user)
+        payload = apply_model_system_prompt_to_body(params, payload, metadata)

        # Check if user has access to the model
        if not bypass_filter and user.role == "user":
@@ -573,6 +587,7 @@ async def generate_chat_completion(
                detail="Model not found",
            )

+    await get_all_models(request)
    model = request.app.state.OPENAI_MODELS.get(model_id)
    if model:
        idx = model["urlIdx"]
@@ -584,7 +599,10 @@ async def generate_chat_completion(

    # Get the API config for the model
    api_config = request.app.state.config.OPENAI_API_CONFIGS.get(
-        request.app.state.config.OPENAI_API_BASE_URLS[idx], {}
+        str(idx),
+        request.app.state.config.OPENAI_API_CONFIGS.get(
+            request.app.state.config.OPENAI_API_BASE_URLS[idx], {}
+        ),  # Legacy support
    )

    prefix_id = api_config.get("prefix_id", None)
@@ -112,7 +112,12 @@ async def update_prompt_by_command(
            detail=ERROR_MESSAGES.NOT_FOUND,
        )

-    if prompt.user_id != user.id and user.role != "admin":
+    # Is the user the original creator, in a group with write access, or an admin
+    if (
+        prompt.user_id != user.id
+        and not has_access(user.id, "write", prompt.access_control)
+        and user.role != "admin"
+    ):
        raise HTTPException(
            status_code=status.HTTP_401_UNAUTHORIZED,
            detail=ERROR_MESSAGES.ACCESS_PROHIBITED,
@@ -142,7 +147,11 @@ async def delete_prompt_by_command(command: str, user=Depends(get_verified_user)
            detail=ERROR_MESSAGES.NOT_FOUND,
        )

-    if prompt.user_id != user.id and user.role != "admin":
+    if (
+        prompt.user_id != user.id
+        and not has_access(user.id, "write", prompt.access_control)
+        and user.role != "admin"
+    ):
        raise HTTPException(
            status_code=status.HTTP_401_UNAUTHORIZED,
            detail=ERROR_MESSAGES.ACCESS_PROHIBITED,
@@ -55,6 +55,7 @@ from open_webui.retrieval.web.serply import search_serply
 from open_webui.retrieval.web.serpstack import search_serpstack
 from open_webui.retrieval.web.tavily import search_tavily
 from open_webui.retrieval.web.bing import search_bing
+from open_webui.retrieval.web.exa import search_exa


 from open_webui.retrieval.utils import (
@@ -384,10 +385,11 @@ async def get_rag_config(request: Request, user=Depends(get_admin_user)):
                "serply_api_key": request.app.state.config.SERPLY_API_KEY,
                "tavily_api_key": request.app.state.config.TAVILY_API_KEY,
                "searchapi_api_key": request.app.state.config.SEARCHAPI_API_KEY,
-                "seaarchapi_engine": request.app.state.config.SEARCHAPI_ENGINE,
+                "searchapi_engine": request.app.state.config.SEARCHAPI_ENGINE,
                "jina_api_key": request.app.state.config.JINA_API_KEY,
                "bing_search_v7_endpoint": request.app.state.config.BING_SEARCH_V7_ENDPOINT,
                "bing_search_v7_subscription_key": request.app.state.config.BING_SEARCH_V7_SUBSCRIPTION_KEY,
+                "exa_api_key": request.app.state.config.EXA_API_KEY,
                "result_count": request.app.state.config.RAG_WEB_SEARCH_RESULT_COUNT,
                "concurrent_requests": request.app.state.config.RAG_WEB_SEARCH_CONCURRENT_REQUESTS,
            },
@@ -436,6 +438,7 @@ class WebSearchConfig(BaseModel):
    jina_api_key: Optional[str] = None
    bing_search_v7_endpoint: Optional[str] = None
    bing_search_v7_subscription_key: Optional[str] = None
+    exa_api_key: Optional[str] = None
    result_count: Optional[int] = None
    concurrent_requests: Optional[int] = None

@@ -542,6 +545,8 @@ async def update_rag_config(
            form_data.web.search.bing_search_v7_subscription_key
        )

+        request.app.state.config.EXA_API_KEY = form_data.web.search.exa_api_key
+
        request.app.state.config.RAG_WEB_SEARCH_RESULT_COUNT = (
            form_data.web.search.result_count
        )
@@ -591,6 +596,7 @@ async def update_rag_config(
                "jina_api_key": request.app.state.config.JINA_API_KEY,
                "bing_search_v7_endpoint": request.app.state.config.BING_SEARCH_V7_ENDPOINT,
                "bing_search_v7_subscription_key": request.app.state.config.BING_SEARCH_V7_SUBSCRIPTION_KEY,
+                "exa_api_key": request.app.state.config.EXA_API_KEY,
                "result_count": request.app.state.config.RAG_WEB_SEARCH_RESULT_COUNT,
                "concurrent_requests": request.app.state.config.RAG_WEB_SEARCH_CONCURRENT_REQUESTS,
            },
@@ -660,6 +666,7 @@ def save_docs_to_vector_db(
    overwrite: bool = False,
    split: bool = True,
    add: bool = False,
+    user=None,
 ) -> bool:
    def _get_docs_info(docs: list[Document]) -> str:
        docs_info = set()
@@ -775,7 +782,7 @@ def save_docs_to_vector_db(
        )

        embeddings = embedding_function(
-            list(map(lambda x: x.replace("\n", " "), texts))
+            list(map(lambda x: x.replace("\n", " "), texts)), user=user
        )

        items = [
@@ -933,6 +940,7 @@ def process_file(
                    "hash": hash,
                },
                add=(True if form_data.collection_name else False),
+                user=user,
            )

            if result:
@@ -990,7 +998,7 @@ def process_text(
    text_content = form_data.content
    log.debug(f"text_content: {text_content}")

-    result = save_docs_to_vector_db(request, docs, collection_name)
+    result = save_docs_to_vector_db(request, docs, collection_name, user=user)
    if result:
        return {
            "status": True,
@@ -1023,7 +1031,9 @@ def process_youtube_video(
        content = " ".join([doc.page_content for doc in docs])
        log.debug(f"text_content: {content}")

-        save_docs_to_vector_db(request, docs, collection_name, overwrite=True)
+        save_docs_to_vector_db(
+            request, docs, collection_name, overwrite=True, user=user
+        )

        return {
            "status": True,
@@ -1064,7 +1074,9 @@ def process_web(
        content = " ".join([doc.page_content for doc in docs])

        log.debug(f"text_content: {content}")
-        save_docs_to_vector_db(request, docs, collection_name, overwrite=True)
+        save_docs_to_vector_db(
+            request, docs, collection_name, overwrite=True, user=user
+        )

        return {
            "status": True,
@@ -1099,6 +1111,7 @@ def search_web(request: Request, engine: str, query: str) -> list[SearchResult]:
    - SERPER_API_KEY
    - SERPLY_API_KEY
    - TAVILY_API_KEY
+    - EXA_API_KEY
    - SEARCHAPI_API_KEY + SEARCHAPI_ENGINE (by default `google`)
    Args:
        query (str): The query to search for
@@ -1233,6 +1246,13 @@ def search_web(request: Request, engine: str, query: str) -> list[SearchResult]:
            request.app.state.config.RAG_WEB_SEARCH_RESULT_COUNT,
            request.app.state.config.RAG_WEB_SEARCH_DOMAIN_FILTER_LIST,
        )
+    elif engine == "exa":
+        return search_exa(
+            request.app.state.config.EXA_API_KEY,
+            query,
+            request.app.state.config.RAG_WEB_SEARCH_RESULT_COUNT,
+            request.app.state.config.RAG_WEB_SEARCH_DOMAIN_FILTER_LIST,
+        )
    else:
        raise Exception("No search engine API key found in environment variables")

@@ -1272,7 +1292,9 @@ def process_web_search(
            requests_per_second=request.app.state.config.RAG_WEB_SEARCH_CONCURRENT_REQUESTS,
        )
        docs = loader.load()
-        save_docs_to_vector_db(request, docs, collection_name, overwrite=True)
+        save_docs_to_vector_db(
+            request, docs, collection_name, overwrite=True, user=user
+        )

        return {
            "status": True,
@@ -1306,7 +1328,9 @@ def query_doc_handler(
            return query_doc_with_hybrid_search(
                collection_name=form_data.collection_name,
                query=form_data.query,
-                embedding_function=request.app.state.EMBEDDING_FUNCTION,
+                embedding_function=lambda query: request.app.state.EMBEDDING_FUNCTION(
+                    query, user=user
+                ),
                k=form_data.k if form_data.k else request.app.state.config.TOP_K,
                reranking_function=request.app.state.rf,
                r=(
@@ -1314,12 +1338,16 @@ def query_doc_handler(
                    if form_data.r
                    else request.app.state.config.RELEVANCE_THRESHOLD
                ),
+                user=user,
            )
        else:
            return query_doc(
                collection_name=form_data.collection_name,
-                query_embedding=request.app.state.EMBEDDING_FUNCTION(form_data.query),
+                query_embedding=request.app.state.EMBEDDING_FUNCTION(
+                    form_data.query, user=user
+                ),
                k=form_data.k if form_data.k else request.app.state.config.TOP_K,
+                user=user,
            )
    except Exception as e:
        log.exception(e)
@@ -1348,7 +1376,9 @@ def query_collection_handler(
            return query_collection_with_hybrid_search(
                collection_names=form_data.collection_names,
                queries=[form_data.query],
-                embedding_function=request.app.state.EMBEDDING_FUNCTION,
+                embedding_function=lambda query: request.app.state.EMBEDDING_FUNCTION(
+                    query, user=user
+                ),
                k=form_data.k if form_data.k else request.app.state.config.TOP_K,
                reranking_function=request.app.state.rf,
                r=(
@@ -1361,7 +1391,9 @@ def query_collection_handler(
            return query_collection(
                collection_names=form_data.collection_names,
                queries=[form_data.query],
-                embedding_function=request.app.state.EMBEDDING_FUNCTION,
+                embedding_function=lambda query: request.app.state.EMBEDDING_FUNCTION(
+                    query, user=user
+                ),
                k=form_data.k if form_data.k else request.app.state.config.TOP_K,
            )

@@ -1458,6 +1490,7 @@ class BatchProcessFilesResponse(BaseModel):

@router.post("/process/files/batch")
 def process_files_batch(
+    request: Request,
    form_data: BatchProcessFilesForm,
    user=Depends(get_verified_user),
 ) -> BatchProcessFilesResponse:
@@ -1504,7 +1537,11 @@ def process_files_batch(
    if all_docs:
        try:
            save_docs_to_vector_db(
-                docs=all_docs, collection_name=collection_name, add=True
+                request=request,
+                docs=all_docs,
+                collection_name=collection_name,
+                add=True,
+                user=user,
            )

            # Update all files with collection name
@@ -4,11 +4,13 @@ from fastapi.responses import JSONResponse, RedirectResponse
 from pydantic import BaseModel
 from typing import Optional
 import logging
+import re

 from open_webui.utils.chat import generate_chat_completion
 from open_webui.utils.task import (
    title_generation_template,
    query_generation_template,
+    image_prompt_generation_template,
    autocomplete_generation_template,
    tags_generation_template,
    emoji_generation_template,
@@ -23,6 +25,7 @@ from open_webui.utils.task import get_task_model_id
 from open_webui.config import (
    DEFAULT_TITLE_GENERATION_PROMPT_TEMPLATE,
    DEFAULT_TAGS_GENERATION_PROMPT_TEMPLATE,
+    DEFAULT_IMAGE_PROMPT_GENERATION_PROMPT_TEMPLATE,
    DEFAULT_QUERY_GENERATION_PROMPT_TEMPLATE,
    DEFAULT_AUTOCOMPLETE_GENERATION_PROMPT_TEMPLATE,
    DEFAULT_EMOJI_GENERATION_PROMPT_TEMPLATE,
@@ -50,6 +53,7 @@ async def get_task_config(request: Request, user=Depends(get_verified_user)):
        "TASK_MODEL": request.app.state.config.TASK_MODEL,
        "TASK_MODEL_EXTERNAL": request.app.state.config.TASK_MODEL_EXTERNAL,
        "TITLE_GENERATION_PROMPT_TEMPLATE": request.app.state.config.TITLE_GENERATION_PROMPT_TEMPLATE,
+        "IMAGE_PROMPT_GENERATION_PROMPT_TEMPLATE": request.app.state.config.IMAGE_PROMPT_GENERATION_PROMPT_TEMPLATE,
        "ENABLE_AUTOCOMPLETE_GENERATION": request.app.state.config.ENABLE_AUTOCOMPLETE_GENERATION,
        "AUTOCOMPLETE_GENERATION_INPUT_MAX_LENGTH": request.app.state.config.AUTOCOMPLETE_GENERATION_INPUT_MAX_LENGTH,
        "TAGS_GENERATION_PROMPT_TEMPLATE": request.app.state.config.TAGS_GENERATION_PROMPT_TEMPLATE,
@@ -65,6 +69,7 @@ class TaskConfigForm(BaseModel):
    TASK_MODEL: Optional[str]
    TASK_MODEL_EXTERNAL: Optional[str]
    TITLE_GENERATION_PROMPT_TEMPLATE: str
+    IMAGE_PROMPT_GENERATION_PROMPT_TEMPLATE: str
    ENABLE_AUTOCOMPLETE_GENERATION: bool
    AUTOCOMPLETE_GENERATION_INPUT_MAX_LENGTH: int
    TAGS_GENERATION_PROMPT_TEMPLATE: str
@@ -85,6 +90,10 @@ async def update_task_config(
        form_data.TITLE_GENERATION_PROMPT_TEMPLATE
    )

+    request.app.state.config.IMAGE_PROMPT_GENERATION_PROMPT_TEMPLATE = (
+        form_data.IMAGE_PROMPT_GENERATION_PROMPT_TEMPLATE
+    )
+
    request.app.state.config.ENABLE_AUTOCOMPLETE_GENERATION = (
        form_data.ENABLE_AUTOCOMPLETE_GENERATION
    )
@@ -114,6 +123,7 @@ async def update_task_config(
        "TASK_MODEL": request.app.state.config.TASK_MODEL,
        "TASK_MODEL_EXTERNAL": request.app.state.config.TASK_MODEL_EXTERNAL,
        "TITLE_GENERATION_PROMPT_TEMPLATE": request.app.state.config.TITLE_GENERATION_PROMPT_TEMPLATE,
+        "IMAGE_PROMPT_GENERATION_PROMPT_TEMPLATE": request.app.state.config.IMAGE_PROMPT_GENERATION_PROMPT_TEMPLATE,
        "ENABLE_AUTOCOMPLETE_GENERATION": request.app.state.config.ENABLE_AUTOCOMPLETE_GENERATION,
        "AUTOCOMPLETE_GENERATION_INPUT_MAX_LENGTH": request.app.state.config.AUTOCOMPLETE_GENERATION_INPUT_MAX_LENGTH,
        "TAGS_GENERATION_PROMPT_TEMPLATE": request.app.state.config.TAGS_GENERATION_PROMPT_TEMPLATE,
@@ -156,9 +166,20 @@ async def generate_title(
    else:
        template = DEFAULT_TITLE_GENERATION_PROMPT_TEMPLATE

+    messages = form_data["messages"]
+
+    # Remove reasoning details from the messages
+    for message in messages:
+        message["content"] = re.sub(
+            r"<details\s+type=\"reasoning\"[^>]*>.*?<\/details>",
+            "",
+            message["content"],
+            flags=re.S,
+        ).strip()
+
    content = title_generation_template(
        template,
-        form_data["messages"],
+        messages,
        {
            "name": user.name,
            "location": user.info.get("location") if user.info else None,
@@ -170,10 +191,10 @@ async def generate_title(
        "messages": [{"role": "user", "content": content}],
        "stream": False,
        **(
-            {"max_tokens": 50}
+            {"max_tokens": 1000}
            if models[task_model_id]["owned_by"] == "ollama"
            else {
-                "max_completion_tokens": 50,
+                "max_completion_tokens": 1000,
            }
        ),
        "metadata": {
@@ -256,6 +277,66 @@ async def generate_chat_tags(
        )


+@router.post("/image_prompt/completions")
+async def generate_image_prompt(
+    request: Request, form_data: dict, user=Depends(get_verified_user)
+):
+    models = request.app.state.MODELS
+
+    model_id = form_data["model"]
+    if model_id not in models:
+        raise HTTPException(
+            status_code=status.HTTP_404_NOT_FOUND,
+            detail="Model not found",
+        )
+
+    # Check if the user has a custom task model
+    # If the user has a custom task model, use that model
+    task_model_id = get_task_model_id(
+        model_id,
+        request.app.state.config.TASK_MODEL,
+        request.app.state.config.TASK_MODEL_EXTERNAL,
+        models,
+    )
+
+    log.debug(
+        f"generating image prompt using model {task_model_id} for user {user.email} "
+    )
+
+    if request.app.state.config.IMAGE_PROMPT_GENERATION_PROMPT_TEMPLATE != "":
+        template = request.app.state.config.IMAGE_PROMPT_GENERATION_PROMPT_TEMPLATE
+    else:
+        template = DEFAULT_IMAGE_PROMPT_GENERATION_PROMPT_TEMPLATE
+
+    content = image_prompt_generation_template(
+        template,
+        form_data["messages"],
+        user={
+            "name": user.name,
+        },
+    )
+
+    payload = {
+        "model": task_model_id,
+        "messages": [{"role": "user", "content": content}],
+        "stream": False,
+        "metadata": {
+            "task": str(TASKS.IMAGE_PROMPT_GENERATION),
+            "task_body": form_data,
+            "chat_id": form_data.get("chat_id", None),
+        },
+    }
+
+    try:
+        return await generate_chat_completion(request, form_data=payload, user=user)
+    except Exception as e:
+        log.error("Exception occurred", exc_info=True)
+        return JSONResponse(
+            status_code=status.HTTP_400_BAD_REQUEST,
+            content={"detail": "An internal error has occurred."},
+        )
+
+
@router.post("/queries/completions")
 async def generate_queries(
    request: Request, form_data: dict, user=Depends(get_verified_user)
@@ -499,8 +580,8 @@ async def generate_moa_response(
        "model": task_model_id,
        "messages": [{"role": "user", "content": content}],
        "stream": form_data.get("stream", False),
-        "chat_id": form_data.get("chat_id", None),
        "metadata": {
+            "chat_id": form_data.get("chat_id", None),
            "task": str(TASKS.MOA_RESPONSE_GENERATION),
            "task_body": form_data,
        },
@@ -70,7 +70,7 @@ async def create_new_tools(
    user=Depends(get_verified_user),
 ):
    if user.role != "admin" and not has_permission(
-        user.id, "workspace.knowledge", request.app.state.config.USER_PERMISSIONS
+        user.id, "workspace.tools", request.app.state.config.USER_PERMISSIONS
    ):
        raise HTTPException(
            status_code=status.HTTP_401_UNAUTHORIZED,
@@ -165,7 +165,12 @@ async def update_tools_by_id(
            detail=ERROR_MESSAGES.NOT_FOUND,
        )

-    if tools.user_id != user.id and user.role != "admin":
+    # Is the user the original creator, in a group with write access, or an admin
+    if (
+        tools.user_id != user.id
+        and not has_access(user.id, "write", tools.access_control)
+        and user.role != "admin"
+    ):
        raise HTTPException(
            status_code=status.HTTP_401_UNAUTHORIZED,
            detail=ERROR_MESSAGES.UNAUTHORIZED,
@@ -222,7 +227,11 @@ async def delete_tools_by_id(
            detail=ERROR_MESSAGES.NOT_FOUND,
        )

-    if tools.user_id != user.id and user.role != "admin":
+    if (
+        tools.user_id != user.id
+        and not has_access(user.id, "write", tools.access_control)
+        and user.role != "admin"
+    ):
        raise HTTPException(
            status_code=status.HTTP_401_UNAUTHORIZED,
            detail=ERROR_MESSAGES.UNAUTHORIZED,
@@ -304,6 +313,17 @@ async def update_tools_valves_by_id(
            status_code=status.HTTP_401_UNAUTHORIZED,
            detail=ERROR_MESSAGES.NOT_FOUND,
        )
+
+    if (
+        tools.user_id != user.id
+        and not has_access(user.id, "write", tools.access_control)
+        and user.role != "admin"
+    ):
+        raise HTTPException(
+            status_code=status.HTTP_400_BAD_REQUEST,
+            detail=ERROR_MESSAGES.ACCESS_PROHIBITED,
+        )
+
    if id in request.app.state.TOOLS:
        tools_module = request.app.state.TOOLS[id]
    else:
@@ -62,27 +62,45 @@ async def get_user_permissisions(user=Depends(get_verified_user)):
 # User Default Permissions
 ############################
 class WorkspacePermissions(BaseModel):
-    models: bool
-    knowledge: bool
-    prompts: bool
-    tools: bool
+    models: bool = False
+    knowledge: bool = False
+    prompts: bool = False
+    tools: bool = False


 class ChatPermissions(BaseModel):
-    file_upload: bool
-    delete: bool
-    edit: bool
-    temporary: bool
+    controls: bool = True
+    file_upload: bool = True
+    delete: bool = True
+    edit: bool = True
+    temporary: bool = True
+
+
+class FeaturesPermissions(BaseModel):
+    web_search: bool = True
+    image_generation: bool = True
+    code_interpreter: bool = True


 class UserPermissions(BaseModel):
    workspace: WorkspacePermissions
    chat: ChatPermissions
+    features: FeaturesPermissions


-@router.get("/default/permissions")
+@router.get("/default/permissions", response_model=UserPermissions)
 async def get_user_permissions(request: Request, user=Depends(get_admin_user)):
-    return request.app.state.config.USER_PERMISSIONS
+    return {
+        "workspace": WorkspacePermissions(
+            **request.app.state.config.USER_PERMISSIONS.get("workspace", {})
+        ),
+        "chat": ChatPermissions(
+            **request.app.state.config.USER_PERMISSIONS.get("chat", {})
+        ),
+        "features": FeaturesPermissions(
+            **request.app.state.config.USER_PERMISSIONS.get("features", {})
+        ),
+    }


@router.post("/default/permissions")
@@ -237,6 +237,7 @@ async def channel_events(sid, data):
            "channel-events",
            {
                "channel_id": data["channel_id"],
+                "message_id": data.get("message_id", None),
                "data": event_data,
                "user": UserNameResponse(**SESSION_POOL[sid]).model_dump(),
            },
@@ -292,11 +293,39 @@ def get_event_emitter(request_info):
                event_data.get("data", {}),
            )

+        if "type" in event_data and event_data["type"] == "message":
+            message = Chats.get_message_by_id_and_message_id(
+                request_info["chat_id"],
+                request_info["message_id"],
+            )
+
+            content = message.get("content", "")
+            content += event_data.get("data", {}).get("content", "")
+
+            Chats.upsert_message_to_chat_by_id_and_message_id(
+                request_info["chat_id"],
+                request_info["message_id"],
+                {
+                    "content": content,
+                },
+            )
+
+        if "type" in event_data and event_data["type"] == "replace":
+            content = event_data.get("data", {}).get("content", "")
+
+            Chats.upsert_message_to_chat_by_id_and_message_id(
+                request_info["chat_id"],
+                request_info["message_id"],
+                {
+                    "content": content,
+                },
+            )
+
    return __event_emitter__


 def get_event_call(request_info):
-    async def __event_call__(event_data):
+    async def __event_caller__(event_data):
        response = await sio.call(
            "chat-events",
            {
@@ -308,7 +337,10 @@ def get_event_call(request_info):
        )
        return response

-    return __event_call__
+    return __event_caller__
+
+
+get_event_caller = get_event_call


 def get_user_id_from_session_pool(sid):
@@ -26,7 +26,7 @@ class RedisLock:

    def release_lock(self):
        lock_value = self.redis.get(self.lock_name)
-        if lock_value and lock_value.decode("utf-8") == self.lock_id:
+        if lock_value and lock_value == self.lock_id:
            self.redis.delete(self.lock_name)


@@ -1,121 +1,73 @@
 import os
+import shutil
+import json
+from abc import ABC, abstractmethod
+from typing import BinaryIO, Tuple
+
 import boto3
 from botocore.exceptions import ClientError
-import shutil
-
-
-from typing import BinaryIO, Tuple, Optional, Union
-
-from open_webui.constants import ERROR_MESSAGES
 from open_webui.config import (
-    STORAGE_PROVIDER,
    S3_ACCESS_KEY_ID,
-    S3_SECRET_ACCESS_KEY,
    S3_BUCKET_NAME,
-    S3_REGION_NAME,
    S3_ENDPOINT_URL,
+    S3_REGION_NAME,
+    S3_SECRET_ACCESS_KEY,
+    GCS_BUCKET_NAME,
+    GOOGLE_APPLICATION_CREDENTIALS_JSON,
+    STORAGE_PROVIDER,
    UPLOAD_DIR,
 )
+from google.cloud import storage
+from google.cloud.exceptions import GoogleCloudError, NotFound
+from open_webui.constants import ERROR_MESSAGES


-import boto3
-from botocore.exceptions import ClientError
-from typing import BinaryIO, Tuple, Optional
+class StorageProvider(ABC):
+    @abstractmethod
+    def get_file(self, file_path: str) -> str:
+        pass
+
+    @abstractmethod
+    def upload_file(self, file: BinaryIO, filename: str) -> Tuple[bytes, str]:
+        pass
+
+    @abstractmethod
+    def delete_all_files(self) -> None:
+        pass
+
+    @abstractmethod
+    def delete_file(self, file_path: str) -> None:
+        pass


-class StorageProvider:
-    def __init__(self, provider: Optional[str] = None):
-        self.storage_provider: str = provider or STORAGE_PROVIDER
-
-        self.s3_client = None
-        self.s3_bucket_name: Optional[str] = None
-
-        if self.storage_provider == "s3":
-            self._initialize_s3()
-
-    def _initialize_s3(self) -> None:
-        """Initializes the S3 client and bucket name if using S3 storage."""
-        self.s3_client = boto3.client(
-            "s3",
-            region_name=S3_REGION_NAME,
-            endpoint_url=S3_ENDPOINT_URL,
-            aws_access_key_id=S3_ACCESS_KEY_ID,
-            aws_secret_access_key=S3_SECRET_ACCESS_KEY,
-        )
-        self.bucket_name = S3_BUCKET_NAME
-
-    def _upload_to_s3(self, file_path: str, filename: str) -> Tuple[bytes, str]:
-        """Handles uploading of the file to S3 storage."""
-        if not self.s3_client:
-            raise RuntimeError("S3 Client is not initialized.")
-
-        try:
-            self.s3_client.upload_file(file_path, self.bucket_name, filename)
-            return (
-                open(file_path, "rb").read(),
-                "s3://" + self.bucket_name + "/" + filename,
-            )
-        except ClientError as e:
-            raise RuntimeError(f"Error uploading file to S3: {e}")
-
-    def _upload_to_local(self, contents: bytes, filename: str) -> Tuple[bytes, str]:
-        """Handles uploading of the file to local storage."""
+class LocalStorageProvider(StorageProvider):
+    @staticmethod
+    def upload_file(file: BinaryIO, filename: str) -> Tuple[bytes, str]:
+        contents = file.read()
+        if not contents:
+            raise ValueError(ERROR_MESSAGES.EMPTY_CONTENT)
        file_path = f"{UPLOAD_DIR}/{filename}"
        with open(file_path, "wb") as f:
            f.write(contents)
        return contents, file_path

-    def _get_file_from_s3(self, file_path: str) -> str:
-        """Handles downloading of the file from S3 storage."""
-        if not self.s3_client:
-            raise RuntimeError("S3 Client is not initialized.")
-
-        try:
-            bucket_name, key = file_path.split("//")[1].split("/")
-            local_file_path = f"{UPLOAD_DIR}/{key}"
-            self.s3_client.download_file(bucket_name, key, local_file_path)
-            return local_file_path
-        except ClientError as e:
-            raise RuntimeError(f"Error downloading file from S3: {e}")
-
-    def _get_file_from_local(self, file_path: str) -> str:
+    @staticmethod
+    def get_file(file_path: str) -> str:
        """Handles downloading of the file from local storage."""
        return file_path

-    def _delete_from_s3(self, filename: str) -> None:
-        """Handles deletion of the file from S3 storage."""
-        if not self.s3_client:
-            raise RuntimeError("S3 Client is not initialized.")
-
-        try:
-            self.s3_client.delete_object(Bucket=self.bucket_name, Key=filename)
-        except ClientError as e:
-            raise RuntimeError(f"Error deleting file from S3: {e}")
-
-    def _delete_from_local(self, filename: str) -> None:
+    @staticmethod
+    def delete_file(file_path: str) -> None:
        """Handles deletion of the file from local storage."""
+        filename = file_path.split("/")[-1]
        file_path = f"{UPLOAD_DIR}/{filename}"
        if os.path.isfile(file_path):
            os.remove(file_path)
        else:
            print(f"File {file_path} not found in local storage.")

-    def _delete_all_from_s3(self) -> None:
-        """Handles deletion of all files from S3 storage."""
-        if not self.s3_client:
-            raise RuntimeError("S3 Client is not initialized.")
-
-        try:
-            response = self.s3_client.list_objects_v2(Bucket=self.bucket_name)
-            if "Contents" in response:
-                for content in response["Contents"]:
-                    self.s3_client.delete_object(
-                        Bucket=self.bucket_name, Key=content["Key"]
-                    )
-        except ClientError as e:
-            raise RuntimeError(f"Error deleting all files from S3: {e}")
-
-    def _delete_all_from_local(self) -> None:
+    @staticmethod
+    def delete_all_files() -> None:
        """Handles deletion of all files from local storage."""
        if os.path.exists(UPLOAD_DIR):
            for filename in os.listdir(UPLOAD_DIR):
@@ -130,38 +82,141 @@ class StorageProvider:
        else:
            print(f"Directory {UPLOAD_DIR} not found in local storage.")

-    def upload_file(self, file: BinaryIO, filename: str) -> Tuple[bytes, str]:
-        """Uploads a file either to S3 or the local file system."""
-        contents = file.read()
-        if not contents:
-            raise ValueError(ERROR_MESSAGES.EMPTY_CONTENT)
-        contents, file_path = self._upload_to_local(contents, filename)

-        if self.storage_provider == "s3":
-            return self._upload_to_s3(file_path, filename)
-        return contents, file_path
+class S3StorageProvider(StorageProvider):
+    def __init__(self):
+        self.s3_client = boto3.client(
+            "s3",
+            region_name=S3_REGION_NAME,
+            endpoint_url=S3_ENDPOINT_URL,
+            aws_access_key_id=S3_ACCESS_KEY_ID,
+            aws_secret_access_key=S3_SECRET_ACCESS_KEY,
+        )
+        self.bucket_name = S3_BUCKET_NAME
+
+    def upload_file(self, file: BinaryIO, filename: str) -> Tuple[bytes, str]:
+        """Handles uploading of the file to S3 storage."""
+        _, file_path = LocalStorageProvider.upload_file(file, filename)
+        try:
+            self.s3_client.upload_file(file_path, self.bucket_name, filename)
+            return (
+                open(file_path, "rb").read(),
+                "s3://" + self.bucket_name + "/" + filename,
+            )
+        except ClientError as e:
+            raise RuntimeError(f"Error uploading file to S3: {e}")

    def get_file(self, file_path: str) -> str:
-        """Downloads a file either from S3 or the local file system and returns the file path."""
-        if self.storage_provider == "s3":
-            return self._get_file_from_s3(file_path)
-        return self._get_file_from_local(file_path)
+        """Handles downloading of the file from S3 storage."""
+        try:
+            bucket_name, key = file_path.split("//")[1].split("/")
+            local_file_path = f"{UPLOAD_DIR}/{key}"
+            self.s3_client.download_file(bucket_name, key, local_file_path)
+            return local_file_path
+        except ClientError as e:
+            raise RuntimeError(f"Error downloading file from S3: {e}")

-    def delete_file(self, filename: str) -> None:
-        """Deletes a file either from S3 or the local file system."""
-        if self.storage_provider == "s3":
-            self._delete_from_s3(filename)
+    def delete_file(self, file_path: str) -> None:
+        """Handles deletion of the file from S3 storage."""
+        filename = file_path.split("/")[-1]
+        try:
+            self.s3_client.delete_object(Bucket=self.bucket_name, Key=filename)
+        except ClientError as e:
+            raise RuntimeError(f"Error deleting file from S3: {e}")

        # Always delete from local storage
-        self._delete_from_local(filename)
+        LocalStorageProvider.delete_file(file_path)

    def delete_all_files(self) -> None:
-        """Deletes all files from the storage."""
-        if self.storage_provider == "s3":
-            self._delete_all_from_s3()
+        """Handles deletion of all files from S3 storage."""
+        try:
+            response = self.s3_client.list_objects_v2(Bucket=self.bucket_name)
+            if "Contents" in response:
+                for content in response["Contents"]:
+                    self.s3_client.delete_object(
+                        Bucket=self.bucket_name, Key=content["Key"]
+                    )
+        except ClientError as e:
+            raise RuntimeError(f"Error deleting all files from S3: {e}")

        # Always delete from local storage
-        self._delete_all_from_local()
+        LocalStorageProvider.delete_all_files()


-Storage = StorageProvider(provider=STORAGE_PROVIDER)
+class GCSStorageProvider(StorageProvider):
+    def __init__(self):
+        self.bucket_name = GCS_BUCKET_NAME
+
+        if GOOGLE_APPLICATION_CREDENTIALS_JSON:
+            self.gcs_client = storage.Client.from_service_account_info(
+                info=json.loads(GOOGLE_APPLICATION_CREDENTIALS_JSON)
+            )
+        else:
+            # if no credentials json is provided, credentials will be picked up from the environment
+            # if running on local environment, credentials would be user credentials
+            # if running on a Compute Engine instance, credentials would be from Google Metadata server
+            self.gcs_client = storage.Client()
+        self.bucket = self.gcs_client.bucket(GCS_BUCKET_NAME)
+
+    def upload_file(self, file: BinaryIO, filename: str) -> Tuple[bytes, str]:
+        """Handles uploading of the file to GCS storage."""
+        contents, file_path = LocalStorageProvider.upload_file(file, filename)
+        try:
+            blob = self.bucket.blob(filename)
+            blob.upload_from_filename(file_path)
+            return contents, "gs://" + self.bucket_name + "/" + filename
+        except GoogleCloudError as e:
+            raise RuntimeError(f"Error uploading file to GCS: {e}")
+
+    def get_file(self, file_path: str) -> str:
+        """Handles downloading of the file from GCS storage."""
+        try:
+            filename = file_path.removeprefix("gs://").split("/")[1]
+            local_file_path = f"{UPLOAD_DIR}/{filename}"
+            blob = self.bucket.get_blob(filename)
+            blob.download_to_filename(local_file_path)
+
+            return local_file_path
+        except NotFound as e:
+            raise RuntimeError(f"Error downloading file from GCS: {e}")
+
+    def delete_file(self, file_path: str) -> None:
+        """Handles deletion of the file from GCS storage."""
+        try:
+            filename = file_path.removeprefix("gs://").split("/")[1]
+            blob = self.bucket.get_blob(filename)
+            blob.delete()
+        except NotFound as e:
+            raise RuntimeError(f"Error deleting file from GCS: {e}")
+
+        # Always delete from local storage
+        LocalStorageProvider.delete_file(file_path)
+
+    def delete_all_files(self) -> None:
+        """Handles deletion of all files from GCS storage."""
+        try:
+            blobs = self.bucket.list_blobs()
+
+            for blob in blobs:
+                blob.delete()
+
+        except NotFound as e:
+            raise RuntimeError(f"Error deleting all files from GCS: {e}")
+
+        # Always delete from local storage
+        LocalStorageProvider.delete_all_files()
+
+
+def get_storage_provider(storage_provider: str):
+    if storage_provider == "local":
+        Storage = LocalStorageProvider()
+    elif storage_provider == "s3":
+        Storage = S3StorageProvider()
+    elif storage_provider == "gcs":
+        Storage = GCSStorageProvider()
+    else:
+        raise RuntimeError(f"Unsupported storage provider: {storage_provider}")
+    return Storage
+
+
+Storage = get_storage_provider(STORAGE_PROVIDER)
@@ -0,0 +1,274 @@
+import io
+import os
+import boto3
+import pytest
+from botocore.exceptions import ClientError
+from moto import mock_aws
+from open_webui.storage import provider
+from gcp_storage_emulator.server import create_server
+from google.cloud import storage
+
+
+def mock_upload_dir(monkeypatch, tmp_path):
+    """Fixture to monkey-patch the UPLOAD_DIR and create a temporary directory."""
+    directory = tmp_path / "uploads"
+    directory.mkdir()
+    monkeypatch.setattr(provider, "UPLOAD_DIR", str(directory))
+    return directory
+
+
+def test_imports():
+    provider.StorageProvider
+    provider.LocalStorageProvider
+    provider.S3StorageProvider
+    provider.GCSStorageProvider
+    provider.Storage
+
+
+def test_get_storage_provider():
+    Storage = provider.get_storage_provider("local")
+    assert isinstance(Storage, provider.LocalStorageProvider)
+    Storage = provider.get_storage_provider("s3")
+    assert isinstance(Storage, provider.S3StorageProvider)
+    Storage = provider.get_storage_provider("gcs")
+    assert isinstance(Storage, provider.GCSStorageProvider)
+    with pytest.raises(RuntimeError):
+        provider.get_storage_provider("invalid")
+
+
+def test_class_instantiation():
+    with pytest.raises(TypeError):
+        provider.StorageProvider()
+    with pytest.raises(TypeError):
+
+        class Test(provider.StorageProvider):
+            pass
+
+        Test()
+    provider.LocalStorageProvider()
+    provider.S3StorageProvider()
+    provider.GCSStorageProvider()
+
+
+class TestLocalStorageProvider:
+    Storage = provider.LocalStorageProvider()
+    file_content = b"test content"
+    file_bytesio = io.BytesIO(file_content)
+    filename = "test.txt"
+    filename_extra = "test_exyta.txt"
+    file_bytesio_empty = io.BytesIO()
+
+    def test_upload_file(self, monkeypatch, tmp_path):
+        upload_dir = mock_upload_dir(monkeypatch, tmp_path)
+        contents, file_path = self.Storage.upload_file(self.file_bytesio, self.filename)
+        assert (upload_dir / self.filename).exists()
+        assert (upload_dir / self.filename).read_bytes() == self.file_content
+        assert contents == self.file_content
+        assert file_path == str(upload_dir / self.filename)
+        with pytest.raises(ValueError):
+            self.Storage.upload_file(self.file_bytesio_empty, self.filename)
+
+    def test_get_file(self, monkeypatch, tmp_path):
+        upload_dir = mock_upload_dir(monkeypatch, tmp_path)
+        file_path = str(upload_dir / self.filename)
+        file_path_return = self.Storage.get_file(file_path)
+        assert file_path == file_path_return
+
+    def test_delete_file(self, monkeypatch, tmp_path):
+        upload_dir = mock_upload_dir(monkeypatch, tmp_path)
+        (upload_dir / self.filename).write_bytes(self.file_content)
+        assert (upload_dir / self.filename).exists()
+        file_path = str(upload_dir / self.filename)
+        self.Storage.delete_file(file_path)
+        assert not (upload_dir / self.filename).exists()
+
+    def test_delete_all_files(self, monkeypatch, tmp_path):
+        upload_dir = mock_upload_dir(monkeypatch, tmp_path)
+        (upload_dir / self.filename).write_bytes(self.file_content)
+        (upload_dir / self.filename_extra).write_bytes(self.file_content)
+        self.Storage.delete_all_files()
+        assert not (upload_dir / self.filename).exists()
+        assert not (upload_dir / self.filename_extra).exists()
+
+
+@mock_aws
+class TestS3StorageProvider:
+
+    def __init__(self):
+        self.Storage = provider.S3StorageProvider()
+        self.Storage.bucket_name = "my-bucket"
+        self.s3_client = boto3.resource("s3", region_name="us-east-1")
+        self.file_content = b"test content"
+        self.filename = "test.txt"
+        self.filename_extra = "test_exyta.txt"
+        self.file_bytesio_empty = io.BytesIO()
+        super().__init__()
+
+    def test_upload_file(self, monkeypatch, tmp_path):
+        upload_dir = mock_upload_dir(monkeypatch, tmp_path)
+        # S3 checks
+        with pytest.raises(Exception):
+            self.Storage.upload_file(io.BytesIO(self.file_content), self.filename)
+        self.s3_client.create_bucket(Bucket=self.Storage.bucket_name)
+        contents, s3_file_path = self.Storage.upload_file(
+            io.BytesIO(self.file_content), self.filename
+        )
+        object = self.s3_client.Object(self.Storage.bucket_name, self.filename)
+        assert self.file_content == object.get()["Body"].read()
+        # local checks
+        assert (upload_dir / self.filename).exists()
+        assert (upload_dir / self.filename).read_bytes() == self.file_content
+        assert contents == self.file_content
+        assert s3_file_path == "s3://" + self.Storage.bucket_name + "/" + self.filename
+        with pytest.raises(ValueError):
+            self.Storage.upload_file(self.file_bytesio_empty, self.filename)
+
+    def test_get_file(self, monkeypatch, tmp_path):
+        upload_dir = mock_upload_dir(monkeypatch, tmp_path)
+        self.s3_client.create_bucket(Bucket=self.Storage.bucket_name)
+        contents, s3_file_path = self.Storage.upload_file(
+            io.BytesIO(self.file_content), self.filename
+        )
+        file_path = self.Storage.get_file(s3_file_path)
+        assert file_path == str(upload_dir / self.filename)
+        assert (upload_dir / self.filename).exists()
+
+    def test_delete_file(self, monkeypatch, tmp_path):
+        upload_dir = mock_upload_dir(monkeypatch, tmp_path)
+        self.s3_client.create_bucket(Bucket=self.Storage.bucket_name)
+        contents, s3_file_path = self.Storage.upload_file(
+            io.BytesIO(self.file_content), self.filename
+        )
+        assert (upload_dir / self.filename).exists()
+        self.Storage.delete_file(s3_file_path)
+        assert not (upload_dir / self.filename).exists()
+        with pytest.raises(ClientError) as exc:
+            self.s3_client.Object(self.Storage.bucket_name, self.filename).load()
+        error = exc.value.response["Error"]
+        assert error["Code"] == "404"
+        assert error["Message"] == "Not Found"
+
+    def test_delete_all_files(self, monkeypatch, tmp_path):
+        upload_dir = mock_upload_dir(monkeypatch, tmp_path)
+        # create 2 files
+        self.s3_client.create_bucket(Bucket=self.Storage.bucket_name)
+        self.Storage.upload_file(io.BytesIO(self.file_content), self.filename)
+        object = self.s3_client.Object(self.Storage.bucket_name, self.filename)
+        assert self.file_content == object.get()["Body"].read()
+        assert (upload_dir / self.filename).exists()
+        assert (upload_dir / self.filename).read_bytes() == self.file_content
+        self.Storage.upload_file(io.BytesIO(self.file_content), self.filename_extra)
+        object = self.s3_client.Object(self.Storage.bucket_name, self.filename_extra)
+        assert self.file_content == object.get()["Body"].read()
+        assert (upload_dir / self.filename).exists()
+        assert (upload_dir / self.filename).read_bytes() == self.file_content
+
+        self.Storage.delete_all_files()
+        assert not (upload_dir / self.filename).exists()
+        with pytest.raises(ClientError) as exc:
+            self.s3_client.Object(self.Storage.bucket_name, self.filename).load()
+        error = exc.value.response["Error"]
+        assert error["Code"] == "404"
+        assert error["Message"] == "Not Found"
+        assert not (upload_dir / self.filename_extra).exists()
+        with pytest.raises(ClientError) as exc:
+            self.s3_client.Object(self.Storage.bucket_name, self.filename_extra).load()
+        error = exc.value.response["Error"]
+        assert error["Code"] == "404"
+        assert error["Message"] == "Not Found"
+
+        self.Storage.delete_all_files()
+        assert not (upload_dir / self.filename).exists()
+        assert not (upload_dir / self.filename_extra).exists()
+
+
+class TestGCSStorageProvider:
+    Storage = provider.GCSStorageProvider()
+    Storage.bucket_name = "my-bucket"
+    file_content = b"test content"
+    filename = "test.txt"
+    filename_extra = "test_exyta.txt"
+    file_bytesio_empty = io.BytesIO()
+
+    @pytest.fixture(scope="class")
+    def setup(self):
+        host, port = "localhost", 9023
+
+        server = create_server(host, port, in_memory=True)
+        server.start()
+        os.environ["STORAGE_EMULATOR_HOST"] = f"http://{host}:{port}"
+
+        gcs_client = storage.Client()
+        bucket = gcs_client.bucket(self.Storage.bucket_name)
+        bucket.create()
+        self.Storage.gcs_client, self.Storage.bucket = gcs_client, bucket
+        yield
+        bucket.delete(force=True)
+        server.stop()
+
+    def test_upload_file(self, monkeypatch, tmp_path, setup):
+        upload_dir = mock_upload_dir(monkeypatch, tmp_path)
+        # catch error if bucket does not exist
+        with pytest.raises(Exception):
+            self.Storage.bucket = monkeypatch(self.Storage, "bucket", None)
+            self.Storage.upload_file(io.BytesIO(self.file_content), self.filename)
+        contents, gcs_file_path = self.Storage.upload_file(
+            io.BytesIO(self.file_content), self.filename
+        )
+        object = self.Storage.bucket.get_blob(self.filename)
+        assert self.file_content == object.download_as_bytes()
+        # local checks
+        assert (upload_dir / self.filename).exists()
+        assert (upload_dir / self.filename).read_bytes() == self.file_content
+        assert contents == self.file_content
+        assert gcs_file_path == "gs://" + self.Storage.bucket_name + "/" + self.filename
+        # test error if file is empty
+        with pytest.raises(ValueError):
+            self.Storage.upload_file(self.file_bytesio_empty, self.filename)
+
+    def test_get_file(self, monkeypatch, tmp_path, setup):
+        upload_dir = mock_upload_dir(monkeypatch, tmp_path)
+        contents, gcs_file_path = self.Storage.upload_file(
+            io.BytesIO(self.file_content), self.filename
+        )
+        file_path = self.Storage.get_file(gcs_file_path)
+        assert file_path == str(upload_dir / self.filename)
+        assert (upload_dir / self.filename).exists()
+
+    def test_delete_file(self, monkeypatch, tmp_path, setup):
+        upload_dir = mock_upload_dir(monkeypatch, tmp_path)
+        contents, gcs_file_path = self.Storage.upload_file(
+            io.BytesIO(self.file_content), self.filename
+        )
+        # ensure that local directory has the uploaded file as well
+        assert (upload_dir / self.filename).exists()
+        assert self.Storage.bucket.get_blob(self.filename).name == self.filename
+        self.Storage.delete_file(gcs_file_path)
+        # check that deleting file from gcs will delete the local file as well
+        assert not (upload_dir / self.filename).exists()
+        assert self.Storage.bucket.get_blob(self.filename) == None
+
+    def test_delete_all_files(self, monkeypatch, tmp_path, setup):
+        upload_dir = mock_upload_dir(monkeypatch, tmp_path)
+        # create 2 files
+        self.Storage.upload_file(io.BytesIO(self.file_content), self.filename)
+        object = self.Storage.bucket.get_blob(self.filename)
+        assert (upload_dir / self.filename).exists()
+        assert (upload_dir / self.filename).read_bytes() == self.file_content
+        assert self.Storage.bucket.get_blob(self.filename).name == self.filename
+        assert self.file_content == object.download_as_bytes()
+        self.Storage.upload_file(io.BytesIO(self.file_content), self.filename_extra)
+        object = self.Storage.bucket.get_blob(self.filename_extra)
+        assert (upload_dir / self.filename_extra).exists()
+        assert (upload_dir / self.filename_extra).read_bytes() == self.file_content
+        assert (
+            self.Storage.bucket.get_blob(self.filename_extra).name
+            == self.filename_extra
+        )
+        assert self.file_content == object.download_as_bytes()
+
+        self.Storage.delete_all_files()
+        assert not (upload_dir / self.filename).exists()
+        assert not (upload_dir / self.filename_extra).exists()
+        assert self.Storage.bucket.get_blob(self.filename) == None
+        assert self.Storage.bucket.get_blob(self.filename_extra) == None
@@ -1,9 +1,30 @@
 from typing import Optional, Union, List, Dict, Any
 from open_webui.models.users import Users, UserModel
 from open_webui.models.groups import Groups
+
+
+from open_webui.config import DEFAULT_USER_PERMISSIONS
 import json


+def fill_missing_permissions(
+    permissions: Dict[str, Any], default_permissions: Dict[str, Any]
+) -> Dict[str, Any]:
+    """
+    Recursively fills in missing properties in the permissions dictionary
+    using the default permissions as a template.
+    """
+    for key, value in default_permissions.items():
+        if key not in permissions:
+            permissions[key] = value
+        elif isinstance(value, dict) and isinstance(
+            permissions[key], dict
+        ):  # Both are nested dictionaries
+            permissions[key] = fill_missing_permissions(permissions[key], value)
+
+    return permissions
+
+
 def get_permissions(
    user_id: str,
    default_permissions: Dict[str, Any],
@@ -27,39 +48,45 @@ def get_permissions(
                if key not in permissions:
                    permissions[key] = value
                else:
-                    permissions[key] = permissions[key] or value
+                    permissions[key] = (
+                        permissions[key] or value
+                    )  # Use the most permissive value (True > False)
        return permissions

    user_groups = Groups.get_groups_by_member_id(user_id)

-    # deep copy default permissions to avoid modifying the original dict
+    # Deep copy default permissions to avoid modifying the original dict
    permissions = json.loads(json.dumps(default_permissions))

+    # Combine permissions from all user groups
    for group in user_groups:
        group_permissions = group.permissions
        permissions = combine_permissions(permissions, group_permissions)

+    # Ensure all fields from default_permissions are present and filled in
+    permissions = fill_missing_permissions(permissions, default_permissions)
+
    return permissions


 def has_permission(
    user_id: str,
    permission_key: str,
-    default_permissions: Dict[str, bool] = {},
+    default_permissions: Dict[str, Any] = {},
 ) -> bool:
    """
    Check if a user has a specific permission by checking the group permissions
-    and falls back to default permissions if not found in any group.
+    and fall back to default permissions if not found in any group.

    Permission keys can be hierarchical and separated by dots ('.').
    """

-    def get_permission(permissions: Dict[str, bool], keys: List[str]) -> bool:
+    def get_permission(permissions: Dict[str, Any], keys: List[str]) -> bool:
        """Traverse permissions dict using a list of keys (from dot-split permission_key)."""
        for key in keys:
            if key not in permissions:
                return False  # If any part of the hierarchy is missing, deny access
-            permissions = permissions[key]  # Go one level deeper
+            permissions = permissions[key]  # Traverse one level deeper

        return bool(permissions)  # Return the boolean at the final level

@@ -73,7 +100,10 @@ def has_permission(
        if get_permission(group_permissions, permission_hierarchy):
            return True

-    # Check default permissions afterwards if the group permissions don't allow it
+    # Check default permissions afterward if the group permissions don't allow it
+    default_permissions = fill_missing_permissions(
+        default_permissions, DEFAULT_USER_PERMISSIONS
+    )
    return get_permission(default_permissions, permission_hierarchy)


@@ -96,10 +96,12 @@ def get_current_user(
                status.HTTP_403_FORBIDDEN, detail=ERROR_MESSAGES.API_KEY_NOT_ALLOWED
            )

-        if request.app.state.ENABLE_API_KEY_ENDPOINT_RESTRICTIONS:
+        if request.app.state.config.ENABLE_API_KEY_ENDPOINT_RESTRICTIONS:
            allowed_paths = [
                path.strip()
-                for path in str(request.app.state.API_KEY_ALLOWED_PATHS).split(",")
+                for path in str(
+                    request.app.state.config.API_KEY_ALLOWED_ENDPOINTS
+                ).split(",")
            ]

            if request.url.path not in allowed_paths:
@@ -157,6 +157,9 @@ async def generate_chat_completion(
        )


+chat_completion = generate_chat_completion
+
+
 async def chat_completed(request: Request, form_data: dict, user: Any):
    if not request.app.state.MODELS:
        await get_all_models(request)
@@ -179,6 +182,7 @@ async def chat_completed(request: Request, form_data: dict, user: Any):
            "chat_id": data["chat_id"],
            "message_id": data["id"],
            "session_id": data["session_id"],
+            "user_id": user.id,
        }
    )

@@ -187,6 +191,7 @@ async def chat_completed(request: Request, form_data: dict, user: Any):
            "chat_id": data["chat_id"],
            "message_id": data["id"],
            "session_id": data["session_id"],
+            "user_id": user.id,
        }
    )

@@ -310,6 +315,7 @@ async def chat_action(request: Request, action_id: str, form_data: dict, user: A
            "chat_id": data["chat_id"],
            "message_id": data["id"],
            "session_id": data["session_id"],
+            "user_id": user.id,
        }
    )
    __event_call__ = get_event_call(
@@ -317,6 +323,7 @@ async def chat_action(request: Request, action_id: str, form_data: dict, user: A
            "chat_id": data["chat_id"],
            "message_id": data["id"],
            "session_id": data["session_id"],
+            "user_id": user.id,
        }
    )

@@ -7,6 +7,18 @@ from pathlib import Path
 from typing import Callable, Optional


+import collections.abc
+
+
+def deep_update(d, u):
+    for k, v in u.items():
+        if isinstance(v, collections.abc.Mapping):
+            d[k] = deep_update(d.get(k, {}), v)
+        else:
+            d[k] = v
+    return d
+
+
 def get_message_list(messages, message_id):
    """
    Reconstructs a list of messages in order up to the specified message_id.
@@ -20,7 +32,7 @@ def get_message_list(messages, message_id):
    current_message = messages.get(message_id)

    if not current_message:
-        return f"Message ID {message_id} not found in the history."
+        return None

    # Reconstruct the chain by following the parentId links
    message_list = []
@@ -131,6 +143,44 @@ def add_or_update_system_message(content: str, messages: list[dict]):
    return messages


+def add_or_update_user_message(content: str, messages: list[dict]):
+    """
+    Adds a new user message at the end of the messages list
+    or updates the existing user message at the end.
+
+    :param msg: The message to be added or appended.
+    :param messages: The list of message dictionaries.
+    :return: The updated list of message dictionaries.
+    """
+
+    if messages and messages[-1].get("role") == "user":
+        messages[-1]["content"] = f"{messages[-1]['content']}\n{content}"
+    else:
+        # Insert at the end
+        messages.append({"role": "user", "content": content})
+
+    return messages
+
+
+def append_or_update_assistant_message(content: str, messages: list[dict]):
+    """
+    Adds a new assistant message at the end of the messages list
+    or updates the existing assistant message at the end.
+
+    :param msg: The message to be added or appended.
+    :param messages: The list of message dictionaries.
+    :return: The updated list of message dictionaries.
+    """
+
+    if messages and messages[-1].get("role") == "assistant":
+        messages[-1]["content"] = f"{messages[-1]['content']}\n{content}"
+    else:
+        # Insert at the end
+        messages.append({"role": "assistant", "content": content})
+
+    return messages
+
+
 def openai_chat_message_template(model: str):
    return {
        "id": f"{model}-{str(uuid.uuid4())}",
@@ -141,13 +191,24 @@ def openai_chat_message_template(model: str):


 def openai_chat_chunk_message_template(
-    model: str, message: Optional[str] = None, usage: Optional[dict] = None
+    model: str,
+    content: Optional[str] = None,
+    tool_calls: Optional[list[dict]] = None,
+    usage: Optional[dict] = None,
 ) -> dict:
    template = openai_chat_message_template(model)
    template["object"] = "chat.completion.chunk"
-    if message:
-        template["choices"][0]["delta"] = {"content": message}
-    else:
+
+    template["choices"][0]["index"] = 0
+    template["choices"][0]["delta"] = {}
+
+    if content:
+        template["choices"][0]["delta"]["content"] = content
+
+    if tool_calls:
+        template["choices"][0]["delta"]["tool_calls"] = tool_calls
+
+    if not content and not tool_calls:
        template["choices"][0]["finish_reason"] = "stop"

    if usage:
@@ -34,8 +34,8 @@ from open_webui.config import (
    JWT_EXPIRES_IN,
    AppConfig,
 )
-from open_webui.constants import ERROR_MESSAGES
-from open_webui.env import WEBUI_SESSION_COOKIE_SAME_SITE, WEBUI_SESSION_COOKIE_SECURE
+from open_webui.constants import ERROR_MESSAGES, WEBHOOK_MESSAGES
+from open_webui.env import WEBUI_AUTH_COOKIE_SAME_SITE, WEBUI_AUTH_COOKIE_SECURE
 from open_webui.utils.misc import parse_duration
 from open_webui.utils.auth import get_password_hash, create_token
 from open_webui.utils.webhook import post_webhook
@@ -63,17 +63,8 @@ auth_manager_config.JWT_EXPIRES_IN = JWT_EXPIRES_IN
 class OAuthManager:
    def __init__(self):
        self.oauth = OAuth()
-        for provider_name, provider_config in OAUTH_PROVIDERS.items():
-            self.oauth.register(
-                name=provider_name,
-                client_id=provider_config["client_id"],
-                client_secret=provider_config["client_secret"],
-                server_metadata_url=provider_config["server_metadata_url"],
-                client_kwargs={
-                    "scope": provider_config["scope"],
-                },
-                redirect_uri=provider_config["redirect_uri"],
-            )
+        for _, provider_config in OAUTH_PROVIDERS.items():
+            provider_config["register"](self.oauth)

    def get_client(self, provider_name):
        return self.oauth.create_client(provider_name)
@@ -91,7 +82,8 @@ class OAuthManager:
            oauth_allowed_roles = auth_manager_config.OAUTH_ALLOWED_ROLES
            oauth_admin_roles = auth_manager_config.OAUTH_ADMIN_ROLES
            oauth_roles = None
-            role = "pending"  # Default/fallback role if no matching roles are found
+            # Default/fallback role if no matching roles are found
+            role = auth_manager_config.DEFAULT_USER_ROLE

            # Next block extracts the roles from the user data, accepting nested claims of any depth
            if oauth_claim and oauth_allowed_roles and oauth_admin_roles:
@@ -200,14 +192,14 @@ class OAuthManager:
        except Exception as e:
            log.warning(f"OAuth callback error: {e}")
            raise HTTPException(400, detail=ERROR_MESSAGES.INVALID_CRED)
-        user_data: UserInfo = token["userinfo"]
+        user_data: UserInfo = token.get("userinfo")
        if not user_data:
            user_data: UserInfo = await client.userinfo(token=token)
        if not user_data:
            log.warning(f"OAuth callback failed, user data is missing: {token}")
            raise HTTPException(400, detail=ERROR_MESSAGES.INVALID_CRED)

-        sub = user_data.get("sub")
+        sub = user_data.get(OAUTH_PROVIDERS[provider].get("sub_claim", "sub"))
        if not sub:
            log.warning(f"OAuth callback failed, sub is missing: {user_data}")
            raise HTTPException(400, detail=ERROR_MESSAGES.INVALID_CRED)
@@ -255,12 +247,20 @@ class OAuthManager:
                    raise HTTPException(400, detail=ERROR_MESSAGES.EMAIL_TAKEN)

                picture_claim = auth_manager_config.OAUTH_PICTURE_CLAIM
-                picture_url = user_data.get(picture_claim, "")
+                picture_url = user_data.get(
+                    picture_claim, OAUTH_PROVIDERS[provider].get("picture_url", "")
+                )
                if picture_url:
                    # Download the profile image into a base64 string
                    try:
+                        access_token = token.get("access_token")
+                        get_kwargs = {}
+                        if access_token:
+                            get_kwargs["headers"] = {
+                                "Authorization": f"Bearer {access_token}",
+                            }
                        async with aiohttp.ClientSession() as session:
-                            async with session.get(picture_url) as resp:
+                            async with session.get(picture_url, **get_kwargs) as resp:
                                picture = await resp.read()
                                base64_encoded_picture = base64.b64encode(
                                    picture
@@ -274,11 +274,16 @@ class OAuthManager:
                        log.error(
                            f"Error downloading profile image '{picture_url}': {e}"
                        )
-                        picture_url = ""
+                        picture_url = "/user.png"
                if not picture_url:
                    picture_url = "/user.png"
+
                username_claim = auth_manager_config.OAUTH_USERNAME_CLAIM

+                name = user_data.get(username_claim)
+                if not isinstance(user, str):
+                    name = email
+
                role = self.get_user_role(None, user_data)

                user = Auths.insert_new_auth(
@@ -286,7 +291,7 @@ class OAuthManager:
                    password=get_password_hash(
                        str(uuid.uuid4())
                    ),  # Random password, not used
-                    name=user_data.get(username_claim, "User"),
+                    name=name,
                    profile_image_url=picture_url,
                    role=role,
                    oauth_sub=provider_sub,
@@ -295,12 +300,10 @@ class OAuthManager:
                if auth_manager_config.WEBHOOK_URL:
                    post_webhook(
                        auth_manager_config.WEBHOOK_URL,
-                        auth_manager_config.WEBHOOK_MESSAGES.USER_SIGNUP(user.name),
+                        WEBHOOK_MESSAGES.USER_SIGNUP(user.name),
                        {
                            "action": "signup",
-                            "message": auth_manager_config.WEBHOOK_MESSAGES.USER_SIGNUP(
-                                user.name
-                            ),
+                            "message": WEBHOOK_MESSAGES.USER_SIGNUP(user.name),
                            "user": user.model_dump_json(exclude_none=True),
                        },
                    )
@@ -314,7 +317,7 @@ class OAuthManager:
            expires_delta=parse_duration(auth_manager_config.JWT_EXPIRES_IN),
        )

-        if auth_manager_config.ENABLE_OAUTH_GROUP_MANAGEMENT:
+        if auth_manager_config.ENABLE_OAUTH_GROUP_MANAGEMENT and user.role != "admin":
            self.update_user_groups(
                user=user,
                user_data=user_data,
@@ -326,8 +329,8 @@ class OAuthManager:
            key="token",
            value=jwt_token,
            httponly=True,  # Ensures the cookie is not accessible via JavaScript
-            samesite=WEBUI_SESSION_COOKIE_SAME_SITE,
-            secure=WEBUI_SESSION_COOKIE_SECURE,
+            samesite=WEBUI_AUTH_COOKIE_SAME_SITE,
+            secure=WEBUI_AUTH_COOKIE_SECURE,
        )

        if ENABLE_OAUTH_SIGNUP.value:
@@ -336,8 +339,8 @@ class OAuthManager:
                key="oauth_id_token",
                value=oauth_id_token,
                httponly=True,
-                samesite=WEBUI_SESSION_COOKIE_SAME_SITE,
-                secure=WEBUI_SESSION_COOKIE_SECURE,
+                samesite=WEBUI_AUTH_COOKIE_SAME_SITE,
+                secure=WEBUI_AUTH_COOKIE_SECURE,
            )
        # Redirect back to the frontend with the JWT token
        redirect_url = f"{request.base_url}auth#token={jwt_token}"
@@ -1,4 +1,4 @@
-from open_webui.utils.task import prompt_template
+from open_webui.utils.task import prompt_variables_template
 from open_webui.utils.misc import (
    add_or_update_system_message,
 )
@@ -7,19 +7,18 @@ from typing import Callable, Optional


 # inplace function: form_data is modified
-def apply_model_system_prompt_to_body(params: dict, form_data: dict, user) -> dict:
+def apply_model_system_prompt_to_body(
+    params: dict, form_data: dict, metadata: Optional[dict] = None
+) -> dict:
    system = params.get("system", None)
    if not system:
        return form_data

-    if user:
-        template_params = {
-            "user_name": user.name,
-            "user_location": user.info.get("location") if user.info else None,
-        }
-    else:
-        template_params = {}
-    system = prompt_template(system, **template_params)
+    if metadata:
+        print("apply_model_system_prompt_to_body: metadata", metadata)
+        variables = metadata.get("variables", {})
+        system = prompt_variables_template(system, variables)
+
    form_data["messages"] = add_or_update_system_message(
        system, form_data.get("messages", [])
    )
@@ -47,6 +46,7 @@ def apply_model_params_to_body_openai(params: dict, form_data: dict) -> dict:
        "top_p": float,
        "max_tokens": int,
        "frequency_penalty": float,
+        "reasoning_effort": str,
        "seed": lambda x: x,
        "stop": lambda x: [bytes(s, "utf-8").decode("unicode_escape") for s in x],
    }
@@ -154,12 +154,19 @@ def convert_payload_openai_to_ollama(openai_payload: dict) -> dict:
    )
    ollama_payload["stream"] = openai_payload.get("stream", False)

+    if "tools" in openai_payload:
+        ollama_payload["tools"] = openai_payload["tools"]
+
    if "format" in openai_payload:
        ollama_payload["format"] = openai_payload["format"]

    # If there are advanced parameters in the payload, format them in Ollama's options field
    ollama_options = {}

+    if openai_payload.get("options"):
+        ollama_payload["options"] = openai_payload["options"]
+        ollama_options = openai_payload["options"]
+
    # Handle parameters which map directly
    for param in ["temperature", "top_p", "seed"]:
        if param in openai_payload:
@@ -183,4 +190,7 @@ def convert_payload_openai_to_ollama(openai_payload: dict) -> dict:
    if ollama_options:
        ollama_payload["options"] = ollama_options

+    if "metadata" in openai_payload:
+        ollama_payload["metadata"] = openai_payload["metadata"]
+
    return ollama_payload
@@ -53,6 +53,7 @@ class PDFGenerator:
        # - https://facelessuser.github.io/pymdown-extensions/usage_notes/
        # html_content = markdown(content, extensions=["pymdownx.extra"])

+        content = content.replace("\n", "<br/>")
        html_message = f"""
            <div>
                <div>
@@ -167,9 +167,14 @@ def load_function_module_by_id(function_id, content=None):

 def install_frontmatter_requirements(requirements):
    if requirements:
-        req_list = [req.strip() for req in requirements.split(",")]
-        for req in req_list:
-            log.info(f"Installing requirement: {req}")
-            subprocess.check_call([sys.executable, "-m", "pip", "install", req])
+        try:
+            req_list = [req.strip() for req in requirements.split(",")]
+            for req in req_list:
+                log.info(f"Installing requirement: {req}")
+                subprocess.check_call([sys.executable, "-m", "pip", "install", req])
+        except Exception as e:
+            log.error(f"Error installing package: {req}")
+            raise e
+
    else:
        log.info("No requirements found in frontmatter.")
@@ -1,4 +1,5 @@
 import json
+from uuid import uuid4
 from open_webui.utils.misc import (
    openai_chat_chunk_message_template,
    openai_chat_completion_message_template,
@@ -9,7 +10,48 @@ def convert_response_ollama_to_openai(ollama_response: dict) -> dict:
    model = ollama_response.get("model", "ollama")
    message_content = ollama_response.get("message", {}).get("content", "")

-    response = openai_chat_completion_message_template(model, message_content)
+    data = ollama_response
+    usage = {
+        "response_token/s": (
+            round(
+                (
+                    (
+                        data.get("eval_count", 0)
+                        / ((data.get("eval_duration", 0) / 10_000_000))
+                    )
+                    * 100
+                ),
+                2,
+            )
+            if data.get("eval_duration", 0) > 0
+            else "N/A"
+        ),
+        "prompt_token/s": (
+            round(
+                (
+                    (
+                        data.get("prompt_eval_count", 0)
+                        / ((data.get("prompt_eval_duration", 0) / 10_000_000))
+                    )
+                    * 100
+                ),
+                2,
+            )
+            if data.get("prompt_eval_duration", 0) > 0
+            else "N/A"
+        ),
+        "total_duration": data.get("total_duration", 0),
+        "load_duration": data.get("load_duration", 0),
+        "prompt_eval_count": data.get("prompt_eval_count", 0),
+        "prompt_eval_duration": data.get("prompt_eval_duration", 0),
+        "eval_count": data.get("eval_count", 0),
+        "eval_duration": data.get("eval_duration", 0),
+        "approximate_total": (lambda s: f"{s // 3600}h{(s % 3600) // 60}m{s % 60}s")(
+            (data.get("total_duration", 0) or 0) // 1_000_000_000
+        ),
+    }
+
+    response = openai_chat_completion_message_template(model, message_content, usage)
    return response


@@ -19,6 +61,23 @@ async def convert_streaming_response_ollama_to_openai(ollama_streaming_response)

        model = data.get("model", "ollama")
        message_content = data.get("message", {}).get("content", "")
+        tool_calls = data.get("message", {}).get("tool_calls", None)
+        openai_tool_calls = None
+
+        if tool_calls:
+            openai_tool_calls = []
+            for tool_call in tool_calls:
+                openai_tool_call = {
+                    "index": tool_call.get("index", 0),
+                    "id": tool_call.get("id", f"call_{str(uuid4())}"),
+                    "type": "function",
+                    "function": {
+                        "name": tool_call.get("function", {}).get("name", ""),
+                        "arguments": f"{tool_call.get('function', {}).get('arguments', {})}",
+                    },
+                }
+                openai_tool_calls.append(openai_tool_call)
+
        done = data.get("done", False)

        usage = None
@@ -29,7 +88,7 @@ async def convert_streaming_response_ollama_to_openai(ollama_streaming_response)
                        (
                            (
                                data.get("eval_count", 0)
-                                / ((data.get("eval_duration", 0) / 1_000_000))
+                                / ((data.get("eval_duration", 0) / 10_000_000))
                            )
                            * 100
                        ),
@@ -43,7 +102,7 @@ async def convert_streaming_response_ollama_to_openai(ollama_streaming_response)
                        (
                            (
                                data.get("prompt_eval_count", 0)
-                                / ((data.get("prompt_eval_duration", 0) / 1_000_000))
+                                / ((data.get("prompt_eval_duration", 0) / 10_000_000))
                            )
                            * 100
                        ),
@@ -64,7 +123,7 @@ async def convert_streaming_response_ollama_to_openai(ollama_streaming_response)
            }

        data = openai_chat_chunk_message_template(
-            model, message_content if not done else None, usage
+            model, message_content if not done else None, openai_tool_calls, usage
        )

        line = f"data: {json.dumps(data)}\n\n"
@@ -32,6 +32,12 @@ def get_task_model_id(
    return task_model_id


+def prompt_variables_template(template: str, variables: dict[str, str]) -> str:
+    for variable, value in variables.items():
+        template = template.replace(variable, value)
+    return template
+
+
 def prompt_template(
    template: str, user_name: Optional[str] = None, user_location: Optional[str] = None
 ) -> str:
@@ -217,6 +223,24 @@ def tags_generation_template(
    return template


+def image_prompt_generation_template(
+    template: str, messages: list[dict], user: Optional[dict] = None
+) -> str:
+    prompt = get_last_user_message(messages)
+    template = replace_prompt_variable(template, prompt)
+    template = replace_messages_variable(template, messages)
+
+    template = prompt_template(
+        template,
+        **(
+            {"user_name": user.get("name"), "user_location": user.get("location")}
+            if user
+            else {}
+        ),
+    )
+    return template
+
+
 def emoji_generation_template(
    template: str, prompt: str, user: Optional[dict] = None
 ) -> str:
@@ -61,6 +61,12 @@ def get_tools(
            )

        for spec in tools.specs:
+            # TODO: Fix hack for OpenAI API
+            # Some times breaks OpenAI but others don't. Leaving the comment
+            for val in spec.get("parameters", {}).get("properties", {}).values():
+                if val["type"] == "str":
+                    val["type"] = "string"
+
            # Remove internal parameters
            spec["parameters"]["properties"] = {
                key: val
@@ -73,6 +79,13 @@ def get_tools(
            # convert to function that takes only model params and inserts custom params
            original_func = getattr(module, function_name)
            callable = apply_extra_params_to_tool_function(original_func, extra_params)
+
+            if callable.__doc__ and callable.__doc__.strip() != "":
+                s = re.split(":(param|return)", callable.__doc__, 1)
+                spec["description"] = s[0]
+            else:
+                spec["description"] = function_name
+
            # TODO: This needs to be a pydantic model
            tool_dict = {
                "toolkit_id": tool_id,
@@ -1,9 +1,9 @@
-fastapi==0.111.0
+fastapi==0.115.7
 uvicorn[standard]==0.30.6
 pydantic==2.9.2
 python-multipart==0.0.18

-Flask==3.0.3
+Flask==3.1.0
 Flask-Cors==5.0.0

 python-socketio==5.11.3
@@ -11,14 +11,14 @@ python-jose==3.3.0
 passlib[bcrypt]==1.7.4

 requests==2.32.3
-aiohttp==3.11.8
+aiohttp==3.11.11
 async-timeout
 aiocache
 aiofiles

 sqlalchemy==2.0.32
 alembic==1.14.0
-peewee==3.17.6
+peewee==3.17.8
 peewee-migrate==1.12.2
 psycopg2-binary==2.9.9
 pgvector==0.3.5
@@ -40,14 +40,15 @@ tiktoken

 langchain==0.3.7
 langchain-community==0.3.7
-langchain-chroma==0.1.4

 fake-useragent==1.5.1
-chromadb==0.5.15
+chromadb==0.6.2
 pymilvus==2.5.0
 qdrant-client~=1.12.0
 opensearch-py==2.7.1

+
+transformers
 sentence-transformers==3.3.1
 colbert-ai==0.2.21
 einops==0.8.0
@@ -55,11 +56,11 @@ einops==0.8.0

 ftfy==6.2.3
 pypdf==4.3.1
-fpdf2==2.7.9
-pymdown-extensions==10.11.2
+fpdf2==2.8.2
+pymdown-extensions==10.14.2
 docx2txt==0.8
 python-pptx==1.0.0
-unstructured==0.15.9
+unstructured==0.16.11
 nltk==3.9.1
 Markdown==3.7
 pypandoc==1.13
@@ -67,19 +68,19 @@ pandas==2.2.3
 openpyxl==3.1.5
 pyxlsb==1.0.10
 xlrd==2.0.1
-validators==0.33.0
+validators==0.34.0
 psutil
 sentencepiece
-soundfile==0.12.1
+soundfile==0.13.1

-opencv-python-headless==4.10.0.84
+opencv-python-headless==4.11.0.86
 rapidocr-onnxruntime==1.3.24
 rank-bm25==0.2.2

 faster-whisper==1.0.3

-PyJWT[crypto]==2.9.0
-authlib==1.3.2
+PyJWT[crypto]==2.10.1
+authlib==1.4.1

 black==24.8.0
 langfuse==2.44.0
@@ -88,7 +89,7 @@ pytube==15.0.0

 extract_msg
 pydub
-duckduckgo-search~=6.3.5
+duckduckgo-search~=7.3.0

 ## Google Drive
 google-api-python-client
@@ -101,6 +102,7 @@ pytest~=8.3.2
 pytest-docker~=3.1.1

 googleapis-common-protos==1.63.2
+google-cloud-storage==2.19.0

 ## LDAP
 ldap3==2.9.1
@@ -16,6 +16,9 @@ For the UI configuration, you can set up the Apache VirtualHost as follows:

    ProxyPass / http://server.com:3000/ nocanon
    ProxyPassReverse / http://server.com:3000/
+    # Needed after 0.5
+    ProxyPass / ws://server.com:3000/ nocanon
+    ProxyPassReverse / ws://server.com:3000/

 </VirtualHost>
 ```
@@ -32,6 +35,9 @@ Enable the site first before you can request SSL:

    ProxyPass / http://server.com:3000/ nocanon
    ProxyPassReverse / http://server.com:3000/
+    # Needed after 0.5
+    ProxyPass / ws://server.com:3000/ nocanon
+    ProxyPassReverse / ws://server.com:3000/

    SSLEngine on
    SSLCertificateFile /etc/ssl/virtualmin/170514456861234/ssl.cert
@@ -1,12 +1,12 @@
 {
 	"name": "open-webui",
-	"version": "0.5.2",
+	"version": "0.5.8",
 	"lockfileVersion": 3,
 	"requires": true,
 	"packages": {
 		"": {
 			"name": "open-webui",
-			"version": "0.5.2",
+			"version": "0.5.8",
 			"dependencies": {
 				"@codemirror/lang-javascript": "^6.2.2",
 				"@codemirror/lang-python": "^6.1.6",
@@ -16,6 +16,7 @@
 				"@mediapipe/tasks-vision": "^0.10.17",
 				"@pyscript/core": "^0.4.32",
 				"@sveltejs/adapter-node": "^2.0.0",
+				"@sveltejs/svelte-virtual-list": "^3.0.1",
 				"@tiptap/core": "^2.10.0",
 				"@tiptap/extension-code-block-lowlight": "^2.10.0",
 				"@tiptap/extension-highlight": "^2.10.0",
@@ -40,7 +41,7 @@
 				"i18next-resources-to-backend": "^1.2.0",
 				"idb": "^7.1.1",
 				"js-sha256": "^0.10.1",
-				"katex": "^0.16.9",
+				"katex": "^0.16.21",
 				"marked": "^9.1.0",
 				"mermaid": "^10.9.3",
 				"paneforge": "^0.0.6",
@@ -55,7 +56,7 @@
 				"prosemirror-schema-list": "^1.4.1",
 				"prosemirror-state": "^1.4.3",
 				"prosemirror-view": "^1.34.3",
-				"pyodide": "^0.26.1",
+				"pyodide": "^0.27.2",
 				"socket.io-client": "^4.2.0",
 				"sortablejs": "^1.15.2",
 				"svelte-sonner": "^0.3.19",
@@ -68,6 +69,7 @@
 				"@sveltejs/adapter-static": "^3.0.2",
 				"@sveltejs/kit": "^2.5.20",
 				"@sveltejs/vite-plugin-svelte": "^3.1.1",
+				"@tailwindcss/container-queries": "^0.1.1",
 				"@tailwindcss/typography": "^0.5.13",
 				"@typescript-eslint/eslint-plugin": "^6.17.0",
 				"@typescript-eslint/parser": "^6.17.0",
@@ -88,7 +90,7 @@
 				"tailwindcss": "^3.3.3",
 				"tslib": "^2.4.1",
 				"typescript": "^5.5.4",
-				"vite": "^5.3.5",
+				"vite": "^5.4.14",
 				"vitest": "^1.6.0"
 			},
 			"engines": {
@@ -2291,6 +2293,12 @@
 				"vite": "^5.0.3 || ^6.0.0"
 			}
 		},
+		"node_modules/@sveltejs/svelte-virtual-list": {
+			"version": "3.0.1",
+			"resolved": "https://registry.npmjs.org/@sveltejs/svelte-virtual-list/-/svelte-virtual-list-3.0.1.tgz",
+			"integrity": "sha512-aF9TptS7NKKS7/TqpsxQBSDJ9Q0XBYzBehCeIC5DzdMEgrJZpIYao9LRLnyyo6SVodpapm2B7FE/Lj+FSA5/SQ==",
+			"license": "LIL"
+		},
 		"node_modules/@sveltejs/vite-plugin-svelte": {
 			"version": "3.1.1",
 			"resolved": "https://registry.npmjs.org/@sveltejs/vite-plugin-svelte/-/vite-plugin-svelte-3.1.1.tgz",
@@ -2336,6 +2344,16 @@
 				"tslib": "^2.4.0"
 			}
 		},
+		"node_modules/@tailwindcss/container-queries": {
+			"version": "0.1.1",
+			"resolved": "https://registry.npmjs.org/@tailwindcss/container-queries/-/container-queries-0.1.1.tgz",
+			"integrity": "sha512-p18dswChx6WnTSaJCSGx6lTmrGzNNvm2FtXmiO6AuA1V4U5REyoqwmT6kgAsIMdjo07QdAfYXHJ4hnMtfHzWgA==",
+			"dev": true,
+			"license": "MIT",
+			"peerDependencies": {
+				"tailwindcss": ">=3.2.0"
+			}
+		},
 		"node_modules/@tailwindcss/typography": {
 			"version": "0.5.13",
 			"resolved": "https://registry.npmjs.org/@tailwindcss/typography/-/typography-0.5.13.tgz",
@@ -7103,13 +7121,14 @@
 			}
 		},
 		"node_modules/katex": {
-			"version": "0.16.10",
-			"resolved": "https://registry.npmjs.org/katex/-/katex-0.16.10.tgz",
-			"integrity": "sha512-ZiqaC04tp2O5utMsl2TEZTXxa6WSC4yo0fv5ML++D3QZv/vx2Mct0mTlRx3O+uUkjfuAgOkzsCmq5MiUEsDDdA==",
+			"version": "0.16.21",
+			"resolved": "https://registry.npmjs.org/katex/-/katex-0.16.21.tgz",
+			"integrity": "sha512-XvqR7FgOHtWupfMiigNzmh+MgUVmDGU2kXZm899ZkPfcuoPuFxyHmXsgATDpFZDAXCI8tvinaVcDo8PIIJSo4A==",
 			"funding": [
 				"https://opencollective.com/katex",
 				"https://github.com/sponsors/katex"
 			],
+			"license": "MIT",
 			"dependencies": {
 				"commander": "^8.3.0"
 			},
@@ -9358,9 +9377,10 @@
 			}
 		},
 		"node_modules/pyodide": {
-			"version": "0.26.1",
-			"resolved": "https://registry.npmjs.org/pyodide/-/pyodide-0.26.1.tgz",
-			"integrity": "sha512-P+Gm88nwZqY7uBgjbQH8CqqU6Ei/rDn7pS1t02sNZsbyLJMyE2OVXjgNuqVT3KqYWnyGREUN0DbBUCJqk8R0ew==",
+			"version": "0.27.2",
+			"resolved": "https://registry.npmjs.org/pyodide/-/pyodide-0.27.2.tgz",
+			"integrity": "sha512-sfA2kiUuQVRpWI4BYnU3sX5PaTTt/xrcVEmRzRcId8DzZXGGtPgCBC0gCqjUTUYSa8ofPaSjXmzESc86yvvCHg==",
+			"license": "Apache-2.0",
 			"dependencies": {
 				"ws": "^8.5.0"
 			},
@@ -11670,9 +11690,10 @@
 			}
 		},
 		"node_modules/vite": {
-			"version": "5.4.6",
-			"resolved": "https://registry.npmjs.org/vite/-/vite-5.4.6.tgz",
-			"integrity": "sha512-IeL5f8OO5nylsgzd9tq4qD2QqI0k2CQLGrWD0rCN0EQJZpBK5vJAx0I+GDkMOXxQX/OfFHMuLIx6ddAxGX/k+Q==",
+			"version": "5.4.14",
+			"resolved": "https://registry.npmjs.org/vite/-/vite-5.4.14.tgz",
+			"integrity": "sha512-EK5cY7Q1D8JNhSaPKVK4pwBFvaTmZxEnoKXLG/U9gmdDcihQGNzFlgIvaxezFR4glP1LsuiedwMBqCXH3wZccA==",
+			"license": "MIT",
 			"dependencies": {
 				"esbuild": "^0.21.3",
 				"postcss": "^8.4.43",
@@ -1,9 +1,10 @@
 {
 	"name": "open-webui",
-	"version": "0.5.2",
+	"version": "0.5.8",
 	"private": true,
 	"scripts": {
 		"dev": "npm run pyodide:fetch && vite dev --host",
+		"dev:5050": "npm run pyodide:fetch && vite dev --port 5050",
 		"build": "npm run pyodide:fetch && vite build",
 		"preview": "vite preview",
 		"check": "svelte-kit sync && svelte-check --tsconfig ./tsconfig.json",
@@ -24,6 +25,7 @@
 		"@sveltejs/adapter-static": "^3.0.2",
 		"@sveltejs/kit": "^2.5.20",
 		"@sveltejs/vite-plugin-svelte": "^3.1.1",
+		"@tailwindcss/container-queries": "^0.1.1",
 		"@tailwindcss/typography": "^0.5.13",
 		"@typescript-eslint/eslint-plugin": "^6.17.0",
 		"@typescript-eslint/parser": "^6.17.0",
@@ -44,13 +46,12 @@
 		"tailwindcss": "^3.3.3",
 		"tslib": "^2.4.1",
 		"typescript": "^5.5.4",
-		"vite": "^5.3.5",
+		"vite": "^5.4.14",
 		"vitest": "^1.6.0"
 	},
 	"type": "module",
 	"dependencies": {
 		"@codemirror/lang-javascript": "^6.2.2",
-		"codemirror-lang-hcl": "^0.0.0-beta.2",
 		"@codemirror/lang-python": "^6.1.6",
 		"@codemirror/language-data": "^6.5.1",
 		"@codemirror/theme-one-dark": "^6.1.2",
@@ -58,6 +59,7 @@
 		"@mediapipe/tasks-vision": "^0.10.17",
 		"@pyscript/core": "^0.4.32",
 		"@sveltejs/adapter-node": "^2.0.0",
+		"@sveltejs/svelte-virtual-list": "^3.0.1",
 		"@tiptap/core": "^2.10.0",
 		"@tiptap/extension-code-block-lowlight": "^2.10.0",
 		"@tiptap/extension-highlight": "^2.10.0",
@@ -69,6 +71,7 @@
 		"async": "^3.2.5",
 		"bits-ui": "^0.19.7",
 		"codemirror": "^6.0.1",
+		"codemirror-lang-hcl": "^0.0.0-beta.2",
 		"crc-32": "^1.2.2",
 		"dayjs": "^1.11.10",
 		"dompurify": "^3.1.6",
@@ -81,7 +84,7 @@
 		"i18next-resources-to-backend": "^1.2.0",
 		"idb": "^7.1.1",
 		"js-sha256": "^0.10.1",
-		"katex": "^0.16.9",
+		"katex": "^0.16.21",
 		"marked": "^9.1.0",
 		"mermaid": "^10.9.3",
 		"paneforge": "^0.0.6",
@@ -96,7 +99,7 @@
 		"prosemirror-schema-list": "^1.4.1",
 		"prosemirror-state": "^1.4.3",
 		"prosemirror-view": "^1.34.3",
-		"pyodide": "^0.26.1",
+		"pyodide": "^0.27.2",
 		"socket.io-client": "^4.2.0",
 		"sortablejs": "^1.15.2",
 		"svelte-sonner": "^0.3.19",
@@ -6,12 +6,12 @@ authors = [
 ]
 license = { file = "LICENSE" }
 dependencies = [
-    "fastapi==0.111.0",
+    "fastapi==0.115.7",
    "uvicorn[standard]==0.30.6",
    "pydantic==2.9.2",
    "python-multipart==0.0.18",

-    "Flask==3.0.3",
+    "Flask==3.1.0",
    "Flask-Cors==5.0.0",

    "python-socketio==5.11.3",
@@ -19,14 +19,14 @@ dependencies = [
    "passlib[bcrypt]==1.7.4",

    "requests==2.32.3",
-    "aiohttp==3.11.8",
+    "aiohttp==3.11.11",
    "async-timeout",
    "aiocache",
    "aiofiles",

    "sqlalchemy==2.0.32",
    "alembic==1.14.0",
-    "peewee==3.17.6",
+    "peewee==3.17.8",
    "peewee-migrate==1.12.2",
    "psycopg2-binary==2.9.9",
    "pgvector==0.3.5",
@@ -47,25 +47,25 @@ dependencies = [

    "langchain==0.3.7",
    "langchain-community==0.3.7",
-    "langchain-chroma==0.1.4",

    "fake-useragent==1.5.1",
-    "chromadb==0.5.15",
+    "chromadb==0.6.2",
    "pymilvus==2.5.0",
    "qdrant-client~=1.12.0",
    "opensearch-py==2.7.1",

+    "transformers",
    "sentence-transformers==3.3.1",
    "colbert-ai==0.2.21",
    "einops==0.8.0",

    "ftfy==6.2.3",
    "pypdf==4.3.1",
-    "fpdf2==2.7.9",
-    "pymdown-extensions==10.11.2",
+    "fpdf2==2.8.2",
+    "pymdown-extensions==10.14.2",
    "docx2txt==0.8",
    "python-pptx==1.0.0",
-    "unstructured==0.15.9",
+    "unstructured==0.16.11",
    "nltk==3.9.1",
    "Markdown==3.7",
    "pypandoc==1.13",
@@ -73,19 +73,19 @@ dependencies = [
    "openpyxl==3.1.5",
    "pyxlsb==1.0.10",
    "xlrd==2.0.1",
-    "validators==0.33.0",
+    "validators==0.34.0",
    "psutil",
    "sentencepiece",
-    "soundfile==0.12.1",
+    "soundfile==0.13.1",

-    "opencv-python-headless==4.10.0.84",
+    "opencv-python-headless==4.11.0.86",
    "rapidocr-onnxruntime==1.3.24",
    "rank-bm25==0.2.2",

    "faster-whisper==1.0.3",

-    "PyJWT[crypto]==2.9.0",
-    "authlib==1.3.2",
+    "PyJWT[crypto]==2.10.1",
+    "authlib==1.4.1",

    "black==24.8.0",
    "langfuse==2.44.0",
@@ -94,15 +94,22 @@ dependencies = [

    "extract_msg",
    "pydub",
-    "duckduckgo-search~=6.3.5",
+    "duckduckgo-search~=7.3.0",
+
+    "google-api-python-client",
+    "google-auth-httplib2",
+    "google-auth-oauthlib",

    "docker~=7.1.0",
    "pytest~=8.3.2",
    "pytest-docker~=3.1.1",
+    "moto[s3]>=5.0.26",

    "googleapis-common-protos==1.63.2",
+    "google-cloud-storage==2.19.0",

-    "ldap3==2.9.1"
+    "ldap3==2.9.1",
+    "gcp-storage-emulator>=2024.8.3",
 ]
 readme = "README.md"
 requires-python = ">= 3.11, < 3.13.0a1"
@@ -151,3 +158,10 @@ exclude = [
    "chroma.sqlite3",
 ]
 force-include = { "CHANGELOG.md" = "open_webui/CHANGELOG.md", build = "open_webui/frontend" }
+
+[tool.codespell]
+# Ref: https://github.com/codespell-project/codespell#using-a-config-file
+skip = '.git*,*.svg,package-lock.json,i18n,*.lock,*.css,*-bundle.js,locales,example-doc.txt,emoji-shortcodes.json'
+check-hidden = true
+# ignore-regex = ''
+ignore-words-list = 'ans'
@@ -9,7 +9,10 @@ const packages = [
 	'scikit-learn',
 	'scipy',
 	'regex',
-	'seaborn'
+	'sympy',
+	'tiktoken',
+	'seaborn',
+	'pytz'
 ];

 import { loadPyodide } from 'pyodide';
@@ -53,11 +53,11 @@ math {
 }

 .markdown-prose {
-	@apply prose dark:prose-invert prose-headings:font-semibold prose-hr:my-4 prose-hr:border-gray-100 prose-hr:dark:border-gray-800 prose-p:my-0 prose-img:my-1 prose-headings:my-1 prose-pre:my-0 prose-table:my-0 prose-blockquote:my-0 prose-ul:-my-0 prose-ol:-my-0 prose-li:-my-0 whitespace-pre-line;
+	@apply prose dark:prose-invert prose-blockquote:border-gray-100 prose-blockquote:dark:border-gray-800 prose-blockquote:border-l-2 prose-blockquote:not-italic prose-blockquote:font-normal  prose-headings:font-semibold prose-hr:my-4 prose-hr:border-gray-100 prose-hr:dark:border-gray-800 prose-p:my-0 prose-img:my-1 prose-headings:my-1 prose-pre:my-0 prose-table:my-0 prose-blockquote:my-0 prose-ul:-my-0 prose-ol:-my-0 prose-li:-my-0 whitespace-pre-line;
 }

 .markdown-prose-xs {
-	@apply text-xs prose dark:prose-invert prose-headings:font-semibold prose-hr:my-0  prose-hr:border-gray-100 prose-hr:dark:border-gray-800 prose-p:my-0 prose-img:my-1 prose-headings:my-1 prose-pre:my-0 prose-table:my-0 prose-blockquote:my-0 prose-ul:-my-0 prose-ol:-my-0 prose-li:-my-0 whitespace-pre-line;
+	@apply text-xs prose dark:prose-invert prose-blockquote:border-gray-100 prose-blockquote:dark:border-gray-800 prose-blockquote:border-l-2 prose-blockquote:not-italic prose-blockquote:font-normal  prose-headings:font-semibold prose-hr:my-0  prose-hr:border-gray-100 prose-hr:dark:border-gray-800 prose-p:my-0 prose-img:my-1 prose-headings:my-1 prose-pre:my-0 prose-table:my-0 prose-blockquote:my-0 prose-ul:-my-0 prose-ol:-my-0 prose-li:-my-0 whitespace-pre-line;
 }

 .markdown a {
@@ -68,6 +68,19 @@ math {
 	font-family: 'Archivo', sans-serif;
 }

+.drag-region {
+	-webkit-app-region: drag;
+}
+
+.drag-region a,
+.drag-region button {
+	-webkit-app-region: no-drag;
+}
+
+.no-drag-region {
+	-webkit-app-region: no-drag;
+}
+
 iframe {
 	@apply rounded-lg;
 }
@@ -102,18 +115,62 @@ li p {

 select {
 	background-image: url("data:image/svg+xml;charset=utf-8,%3Csvg xmlns='http://www.w3.org/2000/svg' fill='none' viewBox='0 0 20 20'%3E%3Cpath stroke='%236B7280' stroke-linecap='round' stroke-linejoin='round' stroke-width='1.5' d='m6 8 4 4 4-4'/%3E%3C/svg%3E");
-	background-position: right 0.5rem center;
+	background-position: right 0rem center;
 	background-repeat: no-repeat;
 	background-size: 1.5em 1.5em;
-	padding-right: 2.5rem;
 	-webkit-print-color-adjust: exact;
 	print-color-adjust: exact;
+	/* padding-right: 2.5rem; */
 	/* for Firefox */
 	-moz-appearance: none;
 	/* for Chrome */
 	-webkit-appearance: none;
 }

+@keyframes shimmer {
+	0% {
+		background-position: 200% 0;
+	}
+	100% {
+		background-position: -200% 0;
+	}
+}
+
+.shimmer {
+	background: linear-gradient(90deg, #9a9b9e 25%, #2a2929 50%, #9a9b9e 75%);
+	background-size: 200% 100%;
+	background-clip: text;
+	-webkit-background-clip: text;
+	-webkit-text-fill-color: transparent;
+	animation: shimmer 4s linear infinite;
+	color: #818286; /* Fallback color */
+}
+
+:global(.dark) .shimmer {
+	background: linear-gradient(90deg, #818286 25%, #eae5e5 50%, #818286 75%);
+	background-size: 200% 100%;
+	background-clip: text;
+	-webkit-background-clip: text;
+	-webkit-text-fill-color: transparent;
+	animation: shimmer 4s linear infinite;
+	color: #a1a3a7; /* Darker fallback color for dark mode */
+}
+
+@keyframes smoothFadeIn {
+	0% {
+		opacity: 0;
+		transform: translateY(-10px);
+	}
+	100% {
+		opacity: 1;
+		transform: translateY(0);
+	}
+}
+
+.status-description {
+	animation: smoothFadeIn 0.2s forwards;
+}
+
 .katex-mathml {
 	display: none;
 }
@@ -1,4 +1,5 @@
 import { WEBUI_API_BASE_URL } from '$lib/constants';
+import { t } from 'i18next';

 type ChannelForm = {
 	name: string;
@@ -207,7 +208,48 @@ export const getChannelMessages = async (
 	return res;
 };

+export const getChannelThreadMessages = async (
+	token: string = '',
+	channel_id: string,
+	message_id: string,
+	skip: number = 0,
+	limit: number = 50
+) => {
+	let error = null;
+
+	const res = await fetch(
+		`${WEBUI_API_BASE_URL}/channels/${channel_id}/messages/${message_id}/thread?skip=${skip}&limit=${limit}`,
+		{
+			method: 'GET',
+			headers: {
+				Accept: 'application/json',
+				'Content-Type': 'application/json',
+				authorization: `Bearer ${token}`
+			}
+		}
+	)
+		.then(async (res) => {
+			if (!res.ok) throw await res.json();
+			return res.json();
+		})
+		.then((json) => {
+			return json;
+		})
+		.catch((err) => {
+			error = err.detail;
+			console.log(err);
+			return null;
+		});
+
+	if (error) {
+		throw error;
+	}
+
+	return res;
+};
+
 type MessageForm = {
+	parent_id?: string;
 	content: string;
 	data?: object;
 	meta?: object;
@@ -285,6 +327,86 @@ export const updateMessage = async (
 	return res;
 };

+export const addReaction = async (
+	token: string = '',
+	channel_id: string,
+	message_id: string,
+	name: string
+) => {
+	let error = null;
+
+	const res = await fetch(
+		`${WEBUI_API_BASE_URL}/channels/${channel_id}/messages/${message_id}/reactions/add`,
+		{
+			method: 'POST',
+			headers: {
+				Accept: 'application/json',
+				'Content-Type': 'application/json',
+				authorization: `Bearer ${token}`
+			},
+			body: JSON.stringify({ name })
+		}
+	)
+		.then(async (res) => {
+			if (!res.ok) throw await res.json();
+			return res.json();
+		})
+		.then((json) => {
+			return json;
+		})
+		.catch((err) => {
+			error = err.detail;
+			console.log(err);
+			return null;
+		});
+
+	if (error) {
+		throw error;
+	}
+
+	return res;
+};
+
+export const removeReaction = async (
+	token: string = '',
+	channel_id: string,
+	message_id: string,
+	name: string
+) => {
+	let error = null;
+
+	const res = await fetch(
+		`${WEBUI_API_BASE_URL}/channels/${channel_id}/messages/${message_id}/reactions/remove`,
+		{
+			method: 'POST',
+			headers: {
+				Accept: 'application/json',
+				'Content-Type': 'application/json',
+				authorization: `Bearer ${token}`
+			},
+			body: JSON.stringify({ name })
+		}
+	)
+		.then(async (res) => {
+			if (!res.ok) throw await res.json();
+			return res.json();
+		})
+		.then((json) => {
+			return json;
+		})
+		.catch((err) => {
+			error = err.detail;
+			console.log(err);
+			return null;
+		});
+
+	if (error) {
+		throw error;
+	}
+
+	return res;
+};
+
 export const deleteMessage = async (token: string = '', channel_id: string, message_id: string) => {
 	let error = null;

@@ -580,10 +580,51 @@ export const toggleChatPinnedStatusById = async (token: string, id: string) => {
 	return res;
 };

-export const cloneChatById = async (token: string, id: string) => {
+export const cloneChatById = async (token: string, id: string, title?: string) => {
 	let error = null;

 	const res = await fetch(`${WEBUI_API_BASE_URL}/chats/${id}/clone`, {
+		method: 'POST',
+		headers: {
+			Accept: 'application/json',
+			'Content-Type': 'application/json',
+			...(token && { authorization: `Bearer ${token}` })
+		},
+		body: JSON.stringify({
+			...(title && { title: title })
+		})
+	})
+		.then(async (res) => {
+			if (!res.ok) throw await res.json();
+			return res.json();
+		})
+		.then((json) => {
+			return json;
+		})
+		.catch((err) => {
+			error = err;
+
+			if ('detail' in err) {
+				error = err.detail;
+			} else {
+				error = err;
+			}
+
+			console.log(err);
+			return null;
+		});
+
+	if (error) {
+		throw error;
+	}
+
+	return res;
+};
+
+export const cloneSharedChatById = async (token: string, id: string) => {
+	let error = null;
+
+	const res = await fetch(`${WEBUI_API_BASE_URL}/chats/${id}/clone/shared`, {
 		method: 'POST',
 		headers: {
 			Accept: 'application/json',
@@ -880,13 +880,14 @@ export const getChangelog = async () => {
 	return res;
 };

-export const getVersionUpdates = async () => {
+export const getVersionUpdates = async (token: string) => {
 	let error = null;

 	const res = await fetch(`${WEBUI_BASE_URL}/api/version/updates`, {
 		method: 'GET',
 		headers: {
-			'Content-Type': 'application/json'
+			'Content-Type': 'application/json',
+			Authorization: `Bearer ${token}`
 		}
 	})
 		.then(async (res) => {
@@ -219,7 +219,7 @@ export const deleteModelById = async (token: string, id: string) => {
 			return json;
 		})
 		.catch((err) => {
-			error = err;
+			error = err.detail;

 			console.log(err);
 			return null;
@@ -360,12 +360,7 @@ export const generateChatCompletion = async (token: string = '', body: object) =
 	return [res, controller];
 };

-export const createModel = async (
-	token: string,
-	tagName: string,
-	content: string,
-	urlIdx: string | null = null
-) => {
+export const createModel = async (token: string, payload: object, urlIdx: string | null = null) => {
 	let error = null;

 	const res = await fetch(
@@ -377,10 +372,7 @@ export const createModel = async (
 				'Content-Type': 'application/json',
 				Authorization: `Bearer ${token}`
 			},
-			body: JSON.stringify({
-				name: tagName,
-				modelfile: content
-			})
+			body: JSON.stringify(payload)
 		}
 	).catch((err) => {
 		error = err;
@@ -322,7 +322,7 @@ export const generateOpenAIChatCompletion = async (
 			return res.json();
 		})
 		.catch((err) => {
-			error = `${err?.detail ?? 'Network Problem'}`;
+			error = `${err?.detail ?? err}`;
 			return null;
 		});

@@ -1,5 +1,5 @@
 <script lang="ts">
-	import { settings } from '$lib/stores';
+	import { settings, playingNotificationSound, isLastActiveTab } from '$lib/stores';
 	import DOMPurify from 'dompurify';

 	import { marked } from 'marked';
@@ -12,9 +12,20 @@
 	export let content: string;

 	onMount(() => {
+		if (!navigator.userActivation.hasBeenActive) {
+			return;
+		}
+
 		if ($settings?.notificationSound ?? true) {
-			const audio = new Audio(`/audio/notification.mp3`);
-			audio.play();
+			if (!$playingNotificationSound && $isLastActiveTab) {
+				playingNotificationSound.set(true);
+
+				const audio = new Audio(`/audio/notification.mp3`);
+				audio.play().finally(() => {
+					// Ensure the global state is reset after the sound finishes
+					playingNotificationSound.set(false);
+				});
+			}
 		}
 	});
 </script>
@@ -61,7 +61,7 @@

 	const shareHandler = async (func) => {
 		const item = await getFunctionById(localStorage.token, func.id).catch((error) => {
-			toast.error(error);
+			toast.error(`${error}`);
 			return null;
 		});

@@ -88,7 +88,7 @@

 	const cloneHandler = async (func) => {
 		const _function = await getFunctionById(localStorage.token, func.id).catch((error) => {
-			toast.error(error);
+			toast.error(`${error}`);
 			return null;
 		});

@@ -104,7 +104,7 @@

 	const exportHandler = async (func) => {
 		const _function = await getFunctionById(localStorage.token, func.id).catch((error) => {
-			toast.error(error);
+			toast.error(`${error}`);
 			return null;
 		});

@@ -118,7 +118,7 @@

 	const deleteHandler = async (func) => {
 		const res = await deleteFunctionById(localStorage.token, func.id).catch((error) => {
-			toast.error(error);
+			toast.error(`${error}`);
 			return null;
 		});

@@ -132,7 +132,7 @@

 	const toggleGlobalHandler = async (func) => {
 		const res = await toggleGlobalById(localStorage.token, func.id).catch((error) => {
-			toast.error(error);
+			toast.error(`${error}`);
 		});

 		if (res) {
@@ -418,7 +418,7 @@
 			class="flex text-xs items-center space-x-1 px-3 py-1.5 rounded-xl bg-gray-50 hover:bg-gray-100 dark:bg-gray-800 dark:hover:bg-gray-700 dark:text-gray-200 transition"
 			on:click={async () => {
 				const _functions = await exportFunctions(localStorage.token).catch((error) => {
-					toast.error(error);
+					toast.error(`${error}`);
 					return null;
 				});

@@ -510,7 +510,7 @@

 			for (const func of _functions) {
 				const res = await createNewFunction(localStorage.token, func).catch((error) => {
-					toast.error(error);
+					toast.error(`${error}`);
 					return null;
 				});
 			}
@@ -51,7 +51,7 @@
 			models = [];
 		} else {
 			const res = await _getModels(localStorage.token).catch((e) => {
-				toast.error(e);
+				toast.error(`${e}`);
 			});

 			if (res) {
@@ -74,7 +74,7 @@
 			}, 100);
 		} else {
 			const res = await _getVoices(localStorage.token).catch((e) => {
-				toast.error(e);
+				toast.error(`${e}`);
 			});

 			if (res) {
@@ -43,9 +43,8 @@

 	const updateOpenAIHandler = async () => {
 		if (ENABLE_OPENAI_API !== null) {
-			OPENAI_API_BASE_URLS = OPENAI_API_BASE_URLS.filter(
-				(url, urlIdx) => OPENAI_API_BASE_URLS.indexOf(url) === urlIdx && url !== ''
-			).map((url) => url.replace(/\/$/, ''));
+			// Remove trailing slashes
+			OPENAI_API_BASE_URLS = OPENAI_API_BASE_URLS.map((url) => url.replace(/\/$/, ''));

 			// Check if API KEYS length is same than API URLS length
 			if (OPENAI_API_KEYS.length !== OPENAI_API_BASE_URLS.length) {
@@ -69,7 +68,7 @@
 				OPENAI_API_KEYS: OPENAI_API_KEYS,
 				OPENAI_API_CONFIGS: OPENAI_API_CONFIGS
 			}).catch((error) => {
-				toast.error(error);
+				toast.error(`${error}`);
 			});

 			if (res) {
@@ -81,24 +80,15 @@

 	const updateOllamaHandler = async () => {
 		if (ENABLE_OLLAMA_API !== null) {
-			// Remove duplicate URLs
-			OLLAMA_BASE_URLS = OLLAMA_BASE_URLS.filter(
-				(url, urlIdx) => OLLAMA_BASE_URLS.indexOf(url) === urlIdx && url !== ''
-			).map((url) => url.replace(/\/$/, ''));
-
-			console.log(OLLAMA_BASE_URLS);
-
-			if (OLLAMA_BASE_URLS.length === 0) {
-				ENABLE_OLLAMA_API = false;
-				toast.info($i18n.t('Ollama API disabled'));
-			}
+			// Remove trailing slashes
+			OLLAMA_BASE_URLS = OLLAMA_BASE_URLS.map((url) => url.replace(/\/$/, ''));

 			const res = await updateOllamaConfig(localStorage.token, {
 				ENABLE_OLLAMA_API: ENABLE_OLLAMA_API,
 				OLLAMA_BASE_URLS: OLLAMA_BASE_URLS,
 				OLLAMA_API_CONFIGS: OLLAMA_API_CONFIGS
 			}).catch((error) => {
-				toast.error(error);
+				toast.error(`${error}`);
 			});

 			if (res) {
@@ -111,14 +101,14 @@
 	const addOpenAIConnectionHandler = async (connection) => {
 		OPENAI_API_BASE_URLS = [...OPENAI_API_BASE_URLS, connection.url];
 		OPENAI_API_KEYS = [...OPENAI_API_KEYS, connection.key];
-		OPENAI_API_CONFIGS[connection.url] = connection.config;
+		OPENAI_API_CONFIGS[OPENAI_API_BASE_URLS.length] = connection.config;

 		await updateOpenAIHandler();
 	};

 	const addOllamaConnectionHandler = async (connection) => {
 		OLLAMA_BASE_URLS = [...OLLAMA_BASE_URLS, connection.url];
-		OLLAMA_API_CONFIGS[connection.url] = connection.config;
+		OLLAMA_API_CONFIGS[OLLAMA_BASE_URLS.length] = connection.config;

 		await updateOllamaHandler();
 	};
@@ -148,15 +138,17 @@
 			OLLAMA_API_CONFIGS = ollamaConfig.OLLAMA_API_CONFIGS;

 			if (ENABLE_OPENAI_API) {
-				for (const url of OPENAI_API_BASE_URLS) {
-					if (!OPENAI_API_CONFIGS[url]) {
-						OPENAI_API_CONFIGS[url] = {};
+				// get url and idx
+				for (const [idx, url] of OPENAI_API_BASE_URLS.entries()) {
+					if (!OPENAI_API_CONFIGS[idx]) {
+						// Legacy support, url as key
+						OPENAI_API_CONFIGS[idx] = OPENAI_API_CONFIGS[url] || {};
 					}
 				}

 				OPENAI_API_BASE_URLS.forEach(async (url, idx) => {
-					OPENAI_API_CONFIGS[url] = OPENAI_API_CONFIGS[url] || {};
-					if (!(OPENAI_API_CONFIGS[url]?.enable ?? true)) {
+					OPENAI_API_CONFIGS[idx] = OPENAI_API_CONFIGS[idx] || {};
+					if (!(OPENAI_API_CONFIGS[idx]?.enable ?? true)) {
 						return;
 					}
 					const res = await getOpenAIModels(localStorage.token, idx);
@@ -167,9 +159,9 @@
 			}

 			if (ENABLE_OLLAMA_API) {
-				for (const url of OLLAMA_BASE_URLS) {
-					if (!OLLAMA_API_CONFIGS[url]) {
-						OLLAMA_API_CONFIGS[url] = {};
+				for (const [idx, url] of OLLAMA_BASE_URLS.entries()) {
+					if (!OLLAMA_API_CONFIGS[idx]) {
+						OLLAMA_API_CONFIGS[idx] = OLLAMA_API_CONFIGS[url] || {};
 					}
 				}
 			}
@@ -242,7 +234,7 @@
 										pipeline={pipelineUrls[url] ? true : false}
 										bind:url
 										bind:key={OPENAI_API_KEYS[idx]}
-										bind:config={OPENAI_API_CONFIGS[url]}
+										bind:config={OPENAI_API_CONFIGS[idx]}
 										onSubmit={() => {
 											updateOpenAIHandler();
 										}}
@@ -251,6 +243,8 @@
 												(url, urlIdx) => idx !== urlIdx
 											);
 											OPENAI_API_KEYS = OPENAI_API_KEYS.filter((key, keyIdx) => idx !== keyIdx);
+
+											delete OPENAI_API_CONFIGS[idx];
 										}}
 									/>
 								{/each}
@@ -301,13 +295,14 @@
 								{#each OLLAMA_BASE_URLS as url, idx}
 									<OllamaConnection
 										bind:url
-										bind:config={OLLAMA_API_CONFIGS[url]}
+										bind:config={OLLAMA_API_CONFIGS[idx]}
 										{idx}
 										onSubmit={() => {
 											updateOllamaHandler();
 										}}
 										onDelete={() => {
 											OLLAMA_BASE_URLS = OLLAMA_BASE_URLS.filter((url, urlIdx) => idx !== urlIdx);
+											delete OLLAMA_API_CONFIGS[idx];
 										}}
 									/>
 								{/each}
@@ -37,7 +37,7 @@

 	const verifyOllamaHandler = async () => {
 		const res = await verifyOllamaConnection(localStorage.token, url, key).catch((error) => {
-			toast.error(error);
+			toast.error(`${error}`);
 		});

 		if (res) {
@@ -47,7 +47,7 @@

 	const verifyOpenAIHandler = async () => {
 		const res = await verifyOpenAIConnection(localStorage.token, url, key).catch((error) => {
-			toast.error(error);
+			toast.error(`${error}`);
 		});

 		if (res) {
@@ -9,6 +9,7 @@
 	import Cog6 from '$lib/components/icons/Cog6.svelte';
 	import Wrench from '$lib/components/icons/Wrench.svelte';
 	import ManageOllamaModal from './ManageOllamaModal.svelte';
+	import ArrowDownTray from '$lib/components/icons/ArrowDownTray.svelte';

 	export let onDelete = () => {};
 	export let onSubmit = () => {};
@@ -70,7 +71,7 @@
 				}}
 				type="button"
 			>
-				<Wrench />
+				<ArrowDownTray />
 			</button>
 		</Tooltip>

@@ -47,7 +47,7 @@
 					reader.onload = async (e) => {
 						const res = await importConfig(localStorage.token, JSON.parse(e.target.result)).catch(
 							(error) => {
-								toast.error(error);
+								toast.error(`${error}`);
 							}
 						);

@@ -132,7 +132,7 @@
 							// exportAllUserChats();

 							downloadDatabase(localStorage.token).catch((error) => {
-								toast.error(error);
+								toast.error(`${error}`);
 							});
 						}}
 					>
@@ -119,7 +119,7 @@
 				url: OpenAIUrl
 			}
 		}).catch(async (error) => {
-			toast.error(error);
+			toast.error(`${error}`);
 			await setEmbeddingConfig();
 			return null;
 		});
@@ -142,7 +142,7 @@
 		const res = await updateRerankingConfig(localStorage.token, {
 			reranking_model: rerankingModel
 		}).catch(async (error) => {
-			toast.error(error);
+			toast.error(`${error}`);
 			await setRerankingConfig();
 			return null;
 		});
@@ -258,7 +258,7 @@
 	bind:show={showResetUploadDirConfirm}
 	on:confirm={async () => {
 		const res = await deleteAllFiles(localStorage.token).catch((error) => {
-			toast.error(error);
+			toast.error(`${error}`);
 			return null;
 		});

@@ -272,7 +272,7 @@
 	bind:show={showResetConfirm}
 	on:confirm={() => {
 		const res = resetVectorDB(localStorage.token).catch((error) => {
-			toast.error(error);
+			toast.error(`${error}`);
 			return null;
 		});

@@ -28,6 +28,7 @@
 		label: '',
 		host: '',
 		port: '',
+		attribute_for_mail: 'mail',
 		attribute_for_username: 'uid',
 		app_dn: '',
 		app_dn_password: '',
@@ -41,7 +42,7 @@
 	const updateLdapServerHandler = async () => {
 		if (!ENABLE_LDAP) return;
 		const res = await updateLdapServer(localStorage.token, LDAP_SERVER).catch((error) => {
-			toast.error(error);
+			toast.error(`${error}`);
 			return null;
 		});
 		if (res) {
@@ -342,6 +343,26 @@
 								/>
 							</div>
 						</div>
+						<div class="flex w-full gap-2">
+							<div class="w-full">
+								<div class=" self-center text-xs font-medium min-w-fit mb-1">
+									{$i18n.t('Attribute for Mail')}
+								</div>
+								<Tooltip
+									content={$i18n.t(
+										'The LDAP attribute that maps to the mail that users use to sign in.'
+									)}
+									placement="top-start"
+								>
+									<input
+										class="w-full bg-transparent outline-none py-0.5"
+										required
+										placeholder={$i18n.t('Example: mail')}
+										bind:value={LDAP_SERVER.attribute_for_mail}
+									/>
+								</Tooltip>
+							</div>
+						</div>
 						<div class="flex w-full gap-2">
 							<div class="w-full">
 								<div class=" self-center text-xs font-medium min-w-fit mb-1">
@@ -99,7 +99,7 @@

 	const getModels = async () => {
 		models = await getImageGenerationModels(localStorage.token).catch((error) => {
-			toast.error(error);
+			toast.error(`${error}`);
 			return null;
 		});
 	};
@@ -107,11 +107,11 @@
 	const updateConfigHandler = async () => {
 		const res = await updateConfig(localStorage.token, config)
 			.catch((error) => {
-				toast.error(error);
+				toast.error(`${error}`);
 				return null;
 			})
 			.catch((error) => {
-				toast.error(error);
+				toast.error(`${error}`);
 				return null;
 			});

@@ -159,13 +159,13 @@
 		}

 		await updateConfig(localStorage.token, config).catch((error) => {
-			toast.error(error);
+			toast.error(`${error}`);
 			loading = false;
 			return null;
 		});

 		await updateImageGenerationConfig(localStorage.token, imageGenerationConfig).catch((error) => {
-			toast.error(error);
+			toast.error(`${error}`);
 			loading = false;
 			return null;
 		});
@@ -178,7 +178,7 @@
 	onMount(async () => {
 		if ($user.role === 'admin') {
 			const res = await getConfig(localStorage.token).catch((error) => {
-				toast.error(error);
+				toast.error(`${error}`);
 				return null;
 			});

@@ -211,7 +211,7 @@
 			});

 			const imageConfigRes = await getImageGenerationConfig(localStorage.token).catch((error) => {
-				toast.error(error);
+				toast.error(`${error}`);
 				return null;
 			});

@@ -234,7 +234,7 @@
 				<div class=" mb-1 text-sm font-medium">{$i18n.t('Image Settings')}</div>

 				<div>
-					<div class=" py-0.5 flex w-full justify-between">
+					<div class=" py-1 flex w-full justify-between">
 						<div class=" self-center text-xs font-medium">
 							{$i18n.t('Image Generation (Experimental)')}
 						</div>
@@ -271,11 +271,20 @@
 					</div>
 				</div>

-				<div class=" py-0.5 flex w-full justify-between">
+				{#if config.enabled}
+					<div class=" py-1 flex w-full justify-between">
+						<div class=" self-center text-xs font-medium">{$i18n.t('Image Prompt Generation')}</div>
+						<div class="px-1">
+							<Switch bind:state={config.prompt_generation} />
+						</div>
+					</div>
+				{/if}
+
+				<div class=" py-1 flex w-full justify-between">
 					<div class=" self-center text-xs font-medium">{$i18n.t('Image Generation Engine')}</div>
 					<div class="flex items-center relative">
 						<select
-							class="w-fit pr-8 rounded px-2 p-1 text-xs bg-transparent outline-none text-right"
+							class=" dark:bg-gray-900 w-fit pr-8 cursor-pointer rounded px-2 p-1 text-xs bg-transparent outline-none text-right"
 							bind:value={config.engine}
 							placeholder={$i18n.t('Select Engine')}
 							on:change={async () => {
@@ -309,7 +318,7 @@
 								on:click={async () => {
 									await updateConfigHandler();
 									const res = await verifyConfigUrl(localStorage.token).catch((error) => {
-										toast.error(error);
+										toast.error(`${error}`);
 										return null;
 									});

@@ -445,7 +454,7 @@
 								on:click={async () => {
 									await updateConfigHandler();
 									const res = await verifyConfigUrl(localStorage.token).catch((error) => {
-										toast.error(error);
+										toast.error(`${error}`);
 										return null;
 									});

@@ -24,13 +24,15 @@
 		TASK_MODEL: '',
 		TASK_MODEL_EXTERNAL: '',
 		TITLE_GENERATION_PROMPT_TEMPLATE: '',
+		IMAGE_PROMPT_GENERATION_PROMPT_TEMPLATE: '',
 		ENABLE_AUTOCOMPLETE_GENERATION: true,
 		AUTOCOMPLETE_GENERATION_INPUT_MAX_LENGTH: -1,
 		TAGS_GENERATION_PROMPT_TEMPLATE: '',
 		ENABLE_TAGS_GENERATION: true,
 		ENABLE_SEARCH_QUERY_GENERATION: true,
 		ENABLE_RETRIEVAL_QUERY_GENERATION: true,
-		QUERY_GENERATION_PROMPT_TEMPLATE: ''
+		QUERY_GENERATION_PROMPT_TEMPLATE: '',
+		TOOLS_FUNCTION_CALLING_PROMPT_TEMPLATE: ''
 	};

 	let promptSuggestions = [];
@@ -140,6 +142,22 @@
 					</Tooltip>
 				</div>

+				<div class="mt-3">
+					<div class=" mb-2.5 text-xs font-medium">{$i18n.t('Image Prompt Generation Prompt')}</div>
+
+					<Tooltip
+						content={$i18n.t('Leave empty to use the default prompt, or enter a custom prompt')}
+						placement="top-start"
+					>
+						<Textarea
+							bind:value={taskConfig.IMAGE_PROMPT_GENERATION_PROMPT_TEMPLATE}
+							placeholder={$i18n.t(
+								'Leave empty to use the default prompt, or enter a custom prompt'
+							)}
+						/>
+					</Tooltip>
+				</div>
+
 				<hr class=" border-gray-50 dark:border-gray-850 my-3" />

 				<div class="my-3 flex w-full items-center justify-between">
@@ -234,6 +252,20 @@
 				</div>
 			</div>

+			<div class="mt-3">
+				<div class=" mb-2.5 text-xs font-medium">{$i18n.t('Tools Function Calling Prompt')}</div>
+
+				<Tooltip
+					content={$i18n.t('Leave empty to use the default prompt, or enter a custom prompt')}
+					placement="top-start"
+				>
+					<Textarea
+						bind:value={taskConfig.TOOLS_FUNCTION_CALLING_PROMPT_TEMPLATE}
+						placeholder={$i18n.t('Leave empty to use the default prompt, or enter a custom prompt')}
+					/>
+				</Tooltip>
+			</div>
+
 			<hr class=" border-gray-50 dark:border-gray-850 my-3" />

 			<div class=" space-y-3 {banners.length > 0 ? ' mb-3' : ''}">
@@ -26,6 +26,9 @@
 	import ConfirmDialog from '$lib/components/common/ConfirmDialog.svelte';
 	import Cog6 from '$lib/components/icons/Cog6.svelte';
 	import ConfigureModelsModal from './Models/ConfigureModelsModal.svelte';
+	import Wrench from '$lib/components/icons/Wrench.svelte';
+	import ArrowDownTray from '$lib/components/icons/ArrowDownTray.svelte';
+	import ManageModelsModal from './Models/ManageModelsModal.svelte';

 	let importFiles;
 	let modelsImportInputElement: HTMLInputElement;
@@ -39,6 +42,7 @@
 	let selectedModelId = null;

 	let showConfigModal = false;
+	let showManageModal = false;

 	$: if (models) {
 		filteredModels = models
@@ -128,7 +132,7 @@
 			await toggleModelById(localStorage.token, model.id);
 		}

-		await init();
+		// await init();
 		_models.set(await getModels(localStorage.token));
 	};

@@ -138,6 +142,7 @@
 </script>

 <ConfigureModelsModal bind:show={showConfigModal} initHandler={init} />
+<ManageModelsModal bind:show={showManageModal} />

 {#if models !== null}
 	{#if selectedModelId === null}
@@ -151,10 +156,22 @@
 					>
 				</div>

-				<div>
-					<Tooltip content={$i18n.t('Configure')}>
+				<div class="flex items-center gap-1.5">
+					<Tooltip content={$i18n.t('Manage Models')}>
 						<button
-							class=" px-2.5 py-1 rounded-full flex gap-1 items-center"
+							class=" p-1 rounded-full flex gap-1 items-center"
+							type="button"
+							on:click={() => {
+								showManageModal = true;
+							}}
+						>
+							<ArrowDownTray />
+						</button>
+					</Tooltip>
+
+					<Tooltip content={$i18n.t('Settings')}>
+						<button
+							class=" p-1 rounded-full flex gap-1 items-center"
 							type="button"
 							on:click={() => {
 								showConfigModal = true;
@@ -33,6 +33,24 @@
 		init();
 	}

+	$: if (selectedModelId) {
+		onModelSelect();
+	}
+
+	const onModelSelect = () => {
+		if (selectedModelId === '') {
+			return;
+		}
+
+		if (defaultModelIds.includes(selectedModelId)) {
+			selectedModelId = '';
+			return;
+		}
+
+		defaultModelIds = [...defaultModelIds, selectedModelId];
+		selectedModelId = '';
+	};
+
 	const init = async () => {
 		config = await getModelsConfig(localStorage.token);

@@ -95,7 +113,7 @@
 	<div>
 		<div class=" flex justify-between dark:text-gray-100 px-5 pt-4 pb-2">
 			<div class=" text-lg font-medium self-center font-primary">
-				{$i18n.t('Configure Models')}
+				{$i18n.t('Settings')}
 			</div>
 			<button
 				class="self-center"
@@ -143,6 +161,24 @@
 									<div class="text-xs text-gray-500">{$i18n.t('Default Models')}</div>
 								</div>

+								<div class="flex items-center -mr-1">
+									<select
+										class="w-full py-1 text-sm rounded-lg bg-transparent {selectedModelId
+											? ''
+											: 'text-gray-500'} placeholder:text-gray-300 dark:placeholder:text-gray-700 outline-none"
+										bind:value={selectedModelId}
+									>
+										<option value="">{$i18n.t('Select a model')}</option>
+										{#each $models as model}
+											<option value={model.id} class="bg-gray-50 dark:bg-gray-700"
+												>{model.name}</option
+											>
+										{/each}
+									</select>
+								</div>
+
+								<!-- <hr class=" border-gray-100 dark:border-gray-700/10 my-2.5 w-full" /> -->
+
 								{#if defaultModelIds.length > 0}
 									<div class="flex flex-col">
 										{#each defaultModelIds as modelId, modelIdx}
@@ -170,44 +206,6 @@
 										{$i18n.t('No models selected')}
 									</div>
 								{/if}
-
-								<hr class=" border-gray-100 dark:border-gray-700/10 my-2.5 w-full" />
-
-								<div class="flex items-center">
-									<select
-										class="w-full py-1 text-sm rounded-lg bg-transparent {selectedModelId
-											? ''
-											: 'text-gray-500'} placeholder:text-gray-300 dark:placeholder:text-gray-700 outline-none"
-										bind:value={selectedModelId}
-									>
-										<option value="">{$i18n.t('Select a model')}</option>
-										{#each $models as model}
-											<option value={model.id} class="bg-gray-50 dark:bg-gray-700"
-												>{model.name}</option
-											>
-										{/each}
-									</select>
-
-									<div>
-										<button
-											type="button"
-											on:click={() => {
-												if (selectedModelId === '') {
-													return;
-												}
-
-												if (defaultModelIds.includes(selectedModelId)) {
-													return;
-												}
-
-												defaultModelIds = [...defaultModelIds, selectedModelId];
-												selectedModelId = '';
-											}}
-										>
-											<Plus className="size-3.5" strokeWidth="2" />
-										</button>
-									</div>
-								</div>
 							</div>
 						</div>

@@ -0,0 +1,26 @@
+<script>
+	import { getContext, onMount } from 'svelte';
+	const i18n = getContext('i18n');
+
+	import ManageOllama from './ManageOllama.svelte';
+
+	export let ollamaConfig = null;
+
+	let selectedUrlIdx = 0;
+</script>
+
+{#if ollamaConfig}
+	<div class="flex-1 mb-2.5 pr-1.5 rounded-lg bg-gray-50 dark:text-gray-300 dark:bg-gray-850">
+		<select
+			class="w-full py-2 px-4 text-sm outline-none bg-transparent"
+			bind:value={selectedUrlIdx}
+			placeholder={$i18n.t('Select an Ollama instance')}
+		>
+			{#each ollamaConfig.OLLAMA_BASE_URLS as url, idx}
+				<option value={idx}>{url}</option>
+			{/each}
+		</select>
+	</div>
+
+	<ManageOllama urlIdx={selectedUrlIdx} />
+{/if}
@@ -0,0 +1,110 @@
+<script>
+	import { toast } from 'svelte-sonner';
+
+	import { createEventDispatcher, getContext, onMount } from 'svelte';
+	const i18n = getContext('i18n');
+	const dispatch = createEventDispatcher();
+
+	import { user } from '$lib/stores';
+
+	import Modal from '$lib/components/common/Modal.svelte';
+	import ManageOllama from './Manage/ManageOllama.svelte';
+	import { getOllamaConfig } from '$lib/apis/ollama';
+	import Spinner from '$lib/components/common/Spinner.svelte';
+	import ManageMultipleOllama from './Manage/ManageMultipleOllama.svelte';
+
+	export let show = false;
+
+	let selected = null;
+	let ollamaConfig = null;
+
+	onMount(async () => {
+		if ($user.role === 'admin') {
+			await Promise.all([
+				(async () => {
+					ollamaConfig = await getOllamaConfig(localStorage.token);
+				})()
+			]);
+
+			if (ollamaConfig) {
+				selected = 'ollama';
+				return;
+			}
+
+			selected = '';
+		}
+	});
+</script>
+
+<Modal size="sm" bind:show>
+	<div>
+		<div class=" flex justify-between dark:text-gray-100 px-5 pt-4">
+			<div class=" text-lg font-medium self-center font-primary">
+				{$i18n.t('Manage Models')}
+			</div>
+			<button
+				class="self-center"
+				on:click={() => {
+					show = false;
+				}}
+			>
+				<svg
+					xmlns="http://www.w3.org/2000/svg"
+					viewBox="0 0 20 20"
+					fill="currentColor"
+					class="w-5 h-5"
+				>
+					<path
+						d="M6.28 5.22a.75.75 0 00-1.06 1.06L8.94 10l-3.72 3.72a.75.75 0 101.06 1.06L10 11.06l3.72 3.72a.75.75 0 101.06-1.06L11.06 10l3.72-3.72a.75.75 0 00-1.06-1.06L10 8.94 6.28 5.22z"
+					/>
+				</svg>
+			</button>
+		</div>
+
+		<div class="flex flex-col md:flex-row w-full px-3 pb-4 md:space-x-4 dark:text-gray-200">
+			<div class=" flex flex-col w-full sm:flex-row sm:justify-center sm:space-x-6">
+				{#if selected === ''}
+					<div class=" py-5 text-gray-400 text-xs">
+						<div>
+							{$i18n.t('No inference engine with management support found')}
+						</div>
+					</div>
+				{:else if selected !== null}
+					<div class=" flex w-full flex-col">
+						<div
+							class="flex gap-1 scrollbar-none overflow-x-auto w-fit text-center text-sm font-medium rounded-full bg-transparent dark:text-gray-200"
+						>
+							<button
+								class="min-w-fit rounded-full p-1.5 {selected === 'ollama'
+									? ''
+									: 'text-gray-300 dark:text-gray-600 hover:text-gray-700 dark:hover:text-white'} transition"
+								on:click={() => {
+									selected = 'ollama';
+								}}>{$i18n.t('Ollama')}</button
+							>
+
+							<!-- <button
+								class="min-w-fit rounded-full p-1.5 {selected === 'llamacpp'
+									? ''
+									: 'text-gray-300 dark:text-gray-600 hover:text-gray-700 dark:hover:text-white'} transition"
+								on:click={() => {
+									selected = 'llamacpp';
+								}}>{$i18n.t('Llama.cpp')}</button
+							> -->
+						</div>
+
+						<div class=" px-1.5 py-1">
+							{#if selected === 'ollama'}
+								<ManageMultipleOllama {ollamaConfig} />
+							{/if}
+						</div>
+					</div>
+				{:else}
+					<div class=" py-5">
+						<Spinner />
+					</div>
+				{/if}
+			</div>
+		</div>
+	</div>
+</Modal>
@@ -57,7 +57,7 @@
 				valves,
 				selectedPipelinesUrlIdx
 			).catch((error) => {
-				toast.error(error);
+				toast.error(`${error}`);
 			});

 			if (res) {
@@ -118,7 +118,7 @@
 			pipelineDownloadUrl,
 			selectedPipelinesUrlIdx
 		).catch((error) => {
-			toast.error(error);
+			toast.error(`${error}`);
 			return null;
 		});

@@ -157,7 +157,7 @@
 		}

 		pipelineFiles = null;
-		const pipelineUploadInputElement = document.getElementById('pipeline-upload-input');
+		const pipelineUploadInputElement = document.getElementById('pipelines-upload-input');

 		if (pipelineUploadInputElement) {
 			pipelineUploadInputElement.value = null;
@@ -172,7 +172,7 @@
 			pipelines[selectedPipelineIdx].id,
 			selectedPipelinesUrlIdx
 		).catch((error) => {
-			toast.error(error);
+			toast.error(`${error}`);
 			return null;
 		});

@@ -25,7 +25,8 @@
 		'duckduckgo',
 		'tavily',
 		'jina',
-		'bing'
+		'bing',
+		'exa'
 	];

 	let youtubeLanguage = 'en';
@@ -261,6 +262,17 @@
 									bind:value={webConfig.search.jina_api_key}
 								/>
 							</div>
+						{:else if webConfig.search.engine === 'exa'}
+							<div>
+								<div class=" self-center text-xs font-medium mb-1">
+									{$i18n.t('Exa API Key')}
+								</div>
+
+								<SensitiveInput
+									placeholder={$i18n.t('Enter Exa API Key')}
+									bind:value={webConfig.search.exa_api_key}
+								/>
+							</div>
 						{:else if webConfig.search.engine === 'bing'}
 							<div>
 								<div class=" self-center text-xs font-medium mb-1">
@@ -53,10 +53,16 @@
 			tools: false
 		},
 		chat: {
+			controls: true,
 			file_upload: true,
 			delete: true,
 			edit: true,
 			temporary: true
+		},
+		features: {
+			web_search: true,
+			image_generation: true,
+			code_interpreter: true
 		}
 	};

@@ -69,7 +75,7 @@

 	const addGroupHandler = async (group) => {
 		const res = await createNewGroup(localStorage.token, group).catch((error) => {
-			toast.error(error);
+			toast.error(`${error}`);
 			return null;
 		});

@@ -84,7 +90,7 @@

 		const res = await updateUserDefaultPermissions(localStorage.token, group.permissions).catch(
 			(error) => {
-				toast.error(error);
+				toast.error(`${error}`);
 				return null;
 			}
 		);
@@ -37,10 +37,16 @@
 			tools: false
 		},
 		chat: {
+			controls: true,
 			file_upload: true,
 			delete: true,
 			edit: true,
 			temporary: true
+		},
+		features: {
+			web_search: true,
+			image_generation: true,
+			code_interpreter: true
 		}
 	};
 	export let userIds = [];
@@ -65,20 +71,8 @@
 		if (group) {
 			name = group.name;
 			description = group.description;
-			permissions = group?.permissions ?? {
-				workspace: {
-					models: false,
-					knowledge: false,
-					prompts: false,
-					tools: false
-				},
-				chat: {
-					file_upload: true,
-					delete: true,
-					edit: true,
-					temporary: true
-				}
-			};
+			permissions = group?.permissions ?? {};
+
 			userIds = group?.user_ids ?? [];
 		}
 	};
@@ -23,7 +23,7 @@

 	const updateHandler = async (_group) => {
 		const res = await updateGroupById(localStorage.token, group.id, _group).catch((error) => {
-			toast.error(error);
+			toast.error(`${error}`);
 			return null;
 		});

@@ -35,7 +35,7 @@

 	const deleteHandler = async () => {
 		const res = await deleteGroupById(localStorage.token, group.id).catch((error) => {
-			toast.error(error);
+			toast.error(`${error}`);
 			return null;
 		});

@@ -1,11 +1,12 @@
 <script lang="ts">
-	import { getContext } from 'svelte';
+	import { getContext, onMount } from 'svelte';
 	const i18n = getContext('i18n');

 	import Switch from '$lib/components/common/Switch.svelte';
 	import Tooltip from '$lib/components/common/Tooltip.svelte';

-	export let permissions = {
+	// Default values for permissions
+	const defaultPermissions = {
 		workspace: {
 			models: false,
 			knowledge: false,
@@ -13,12 +14,39 @@
 			tools: false
 		},
 		chat: {
+			controls: true,
 			delete: true,
 			edit: true,
 			temporary: true,
 			file_upload: true
+		},
+		features: {
+			web_search: true,
+			image_generation: true,
+			code_interpreter: true
 		}
 	};
+
+	export let permissions = {};
+
+	// Reactive statement to ensure all fields are present in `permissions`
+	$: {
+		permissions = fillMissingProperties(permissions, defaultPermissions);
+	}
+
+	function fillMissingProperties(obj: any, defaults: any) {
+		return {
+			...defaults,
+			...obj,
+			workspace: { ...defaults.workspace, ...obj.workspace },
+			chat: { ...defaults.chat, ...obj.chat },
+			features: { ...defaults.features, ...obj.features }
+		};
+	}
+
+	onMount(() => {
+		permissions = fillMissingProperties(permissions, defaultPermissions);
+	});
 </script>

 <div>
@@ -169,6 +197,14 @@
 	<div>
 		<div class=" mb-2 text-sm font-medium">{$i18n.t('Chat Permissions')}</div>

+		<div class="  flex w-full justify-between my-2 pr-2">
+			<div class=" self-center text-xs font-medium">
+				{$i18n.t('Allow Chat Controls')}
+			</div>
+
+			<Switch bind:state={permissions.chat.controls} />
+		</div>
+
 		<div class="  flex w-full justify-between my-2 pr-2">
 			<div class=" self-center text-xs font-medium">
 				{$i18n.t('Allow File Upload')}
@@ -201,4 +237,34 @@
 			<Switch bind:state={permissions.chat.temporary} />
 		</div>
 	</div>
+
+	<hr class=" border-gray-50 dark:border-gray-850 my-2" />
+
+	<div>
+		<div class=" mb-2 text-sm font-medium">{$i18n.t('Features Permissions')}</div>
+
+		<div class="  flex w-full justify-between my-2 pr-2">
+			<div class=" self-center text-xs font-medium">
+				{$i18n.t('Web Search')}
+			</div>
+
+			<Switch bind:state={permissions.features.web_search} />
+		</div>
+
+		<div class="  flex w-full justify-between my-2 pr-2">
+			<div class=" self-center text-xs font-medium">
+				{$i18n.t('Image Generation')}
+			</div>
+
+			<Switch bind:state={permissions.features.image_generation} />
+		</div>
+
+		<div class="  flex w-full justify-between my-2 pr-2">
+			<div class=" self-center text-xs font-medium">
+				{$i18n.t('Code Interpreter')}
+			</div>
+
+			<Switch bind:state={permissions.features.code_interpreter} />
+		</div>
+	</div>
 </div>
@@ -6,7 +6,9 @@

 	import dayjs from 'dayjs';
 	import relativeTime from 'dayjs/plugin/relativeTime';
+	import localizedFormat from 'dayjs/plugin/localizedFormat';
 	dayjs.extend(relativeTime);
+	dayjs.extend(localizedFormat);

 	import { toast } from 'svelte-sonner';

@@ -44,7 +46,7 @@

 	const updateRoleHandler = async (id, role) => {
 		const res = await updateUserRole(localStorage.token, id, role).catch((error) => {
-			toast.error(error);
+			toast.error(`${error}`);
 			return null;
 		});

@@ -55,7 +57,7 @@

 	const deleteUserHandler = async (id) => {
 		const res = await deleteUserById(localStorage.token, id).catch((error) => {
-			toast.error(error);
+			toast.error(`${error}`);
 			return null;
 		});
 		if (res) {
@@ -364,7 +366,7 @@
 					</td>

 					<td class=" px-3 py-1">
-						{dayjs(user.created_at * 1000).format($i18n.t('MMMM DD, YYYY'))}
+						{dayjs(user.created_at * 1000).format('LL')}
 					</td>

 					<td class=" px-3 py-1"> {user.oauth_sub ?? ''} </td>
@@ -49,7 +49,7 @@
 				_user.password,
 				_user.role
 			).catch((error) => {
-				toast.error(error);
+				toast.error(`${error}`);
 			});

 			if (res) {
@@ -114,6 +114,8 @@
 				toast.error($i18n.t('File not found.'));
 			}
 		}
+
+		loading = false;
 	};
 </script>

@@ -148,9 +150,13 @@
 						submitHandler();
 					}}
 				>
-					<div class="flex text-center text-sm font-medium rounded-full bg-transparent/10 p-1 mb-2">
+					<div
+						class="flex -mt-2 mb-1.5 gap-1 scrollbar-none overflow-x-auto w-fit text-center text-sm font-medium rounded-full bg-transparent dark:text-gray-200"
+					>
 						<button
-							class="w-full rounded-full p-1.5 {tab === '' ? 'bg-gray-50 dark:bg-gray-850' : ''}"
+							class="min-w-fit rounded-full p-1.5 {tab === ''
+								? ''
+								: 'text-gray-300 dark:text-gray-600 hover:text-gray-700 dark:hover:text-white'} transition"
 							type="button"
 							on:click={() => {
 								tab = '';
@@ -158,23 +164,24 @@
 						>

 						<button
-							class="w-full rounded-full p-1 {tab === 'import'
-								? 'bg-gray-50 dark:bg-gray-850'
-								: ''}"
+							class="min-w-fit rounded-full p-1.5 {tab === 'import'
+								? ''
+								: 'text-gray-300 dark:text-gray-600 hover:text-gray-700 dark:hover:text-white'} transition"
 							type="button"
 							on:click={() => {
 								tab = 'import';
 							}}>{$i18n.t('CSV Import')}</button
 						>
 					</div>
+
 					<div class="px-1">
 						{#if tab === ''}
-							<div class="flex flex-col w-full">
+							<div class="flex flex-col w-full mb-3">
 								<div class=" mb-1 text-xs text-gray-500">{$i18n.t('Role')}</div>

 								<div class="flex-1">
 									<select
-										class="w-full capitalize rounded-lg py-2 px-4 text-sm dark:text-gray-300 dark:bg-gray-850 disabled:text-gray-500 dark:disabled:text-gray-500 outline-none"
+										class="w-full capitalize rounded-lg text-sm bg-transparent dark:disabled:text-gray-500 outline-none"
 										bind:value={_user.role}
 										placeholder={$i18n.t('Enter Your Role')}
 										required
@@ -191,7 +198,7 @@

 								<div class="flex-1">
 									<input
-										class="w-full rounded-lg py-2 px-4 text-sm dark:text-gray-300 dark:bg-gray-850 disabled:text-gray-500 dark:disabled:text-gray-500 outline-none"
+										class="w-full text-sm bg-transparent disabled:text-gray-500 dark:disabled:text-gray-500 outline-none"
 										type="text"
 										bind:value={_user.name}
 										placeholder={$i18n.t('Enter Your Full Name')}
@@ -208,7 +215,7 @@

 								<div class="flex-1">
 									<input
-										class="w-full rounded-lg py-2 px-4 text-sm dark:text-gray-300 dark:bg-gray-850 disabled:text-gray-500 dark:disabled:text-gray-500 outline-none"
+										class="w-full text-sm bg-transparent disabled:text-gray-500 dark:disabled:text-gray-500 outline-none"
 										type="email"
 										bind:value={_user.email}
 										placeholder={$i18n.t('Enter Your Email')}
@@ -222,7 +229,7 @@

 								<div class="flex-1">
 									<input
-										class="w-full rounded-lg py-2 px-4 text-sm dark:text-gray-300 dark:bg-gray-850 disabled:text-gray-500 dark:disabled:text-gray-500 outline-none"
+										class="w-full text-sm bg-transparent disabled:text-gray-500 dark:disabled:text-gray-500 outline-none"
 										type="password"
 										bind:value={_user.password}
 										placeholder={$i18n.t('Enter Your Password')}
@@ -7,9 +7,11 @@
 	import { updateUserById } from '$lib/apis/users';

 	import Modal from '$lib/components/common/Modal.svelte';
+	import localizedFormat from 'dayjs/plugin/localizedFormat';

 	const i18n = getContext('i18n');
 	const dispatch = createEventDispatcher();
+	dayjs.extend(localizedFormat);

 	export let show = false;
 	export let selectedUser;
@@ -24,7 +26,7 @@

 	const submitHandler = async () => {
 		const res = await updateUserById(localStorage.token, selectedUser.id, _user).catch((error) => {
-			toast.error(error);
+			toast.error(`${error}`);
 		});

 		if (res) {
@@ -87,7 +89,7 @@

 							<div class="text-xs text-gray-500">
 								{$i18n.t('Created at')}
-								{dayjs(selectedUser.created_at * 1000).format($i18n.t('MMMM DD, YYYY'))}
+								{dayjs(selectedUser.created_at * 1000).format('LL')}
 							</div>
 						</div>
 					</div>
--- a/Show More
+++ b/Show More