Skip to main content

Overview

The Open WebUI chat interface provides a powerful and intuitive environment for interacting with AI models. It supports multiple models simultaneously, persistent conversations, and rich media interactions.

Core Chat Features

Multi-Model Conversations

Engage with multiple AI models in parallel within a single chat session.
1

Select Models

Choose one or more models from the model selector in the chat interface
2

Send Your Message

Type your message and it will be sent to all selected models simultaneously
3

Compare Responses

View and compare responses from different models side-by-side
Multi-model conversations allow you to leverage the unique strengths of different models for optimal responses.

Conversation Management

Organize and manage your chat history effectively.
Access all your conversations from the sidebar:
  • Pinned Chats: Keep important conversations at the top
  • Recent Chats: Browse by last updated time
  • Archived Chats: Move old conversations out of view
  • Folder Organization: Group related chats together

Chat Operations

Creating Chats

Start new conversations using the API or UI:
# API endpoint: POST /api/v1/chats/new
{
  "chat": {
    "title": "My Conversation",
    "history": {
      "messages": {},
      "currentId": null
    }
  }
}

Updating Messages

Edit messages in existing conversations:
# API endpoint: POST /api/v1/chats/{id}/messages/{message_id}
{
  "content": "Updated message content"
}
Message updates trigger real-time events through WebSocket, keeping all connected clients synchronized.

Tagging System

Organize conversations with custom tags.

Add Tags

Apply tags to categorize your chats by topic, project, or any custom criteria.

Search by Tag

Use tag:tagname in the search box to filter conversations by specific tags.

Auto-Cleanup

Tags are automatically removed when no chats reference them, keeping your tag list clean.

Multiple Tags

Add multiple tags to a single chat for flexible organization.

Markdown & LaTeX Support

Open WebUI supports full Markdown and LaTeX rendering for rich text formatting.
Standard Markdown syntax is fully supported:
  • Bold, italic, and strikethrough text
  • Code blocks with syntax highlighting
  • Lists, tables, and blockquotes
  • Links and images

Chat Statistics

Track detailed metrics about your conversations.

Usage Analytics

View comprehensive statistics for each chat:
  • Message Count: Total messages in current conversation path
  • History Message Count: All messages including branches
  • Model Usage: Breakdown of which models were used
  • Average Response Time: How quickly models respond
  • Message Length: Average content length for user and assistant messages
// Example stats response from routers/chats.py:179
{
  "id": "chat-id",
  "models": {"gpt-4": 5, "claude-2": 3},
  "message_count": 16,
  "history_message_count": 24,
  "average_response_time": 2.5,
  "average_user_message_content_length": 150,
  "average_assistant_message_content_length": 450
}

Export Capabilities

1

Request Export

Navigate to Settings → Data Export or use the API endpoint:
GET /api/v1/chats/stats/export?stream=true
2

Choose Format

  • Paginated JSON: Standard API response with pagination
  • Streaming JSONL: Line-delimited JSON stream for large datasets
3

Filter by Date

Add updated_at parameter to export only recent chats:
GET /api/v1/chats/stats/export?updated_at=1609459200
Export features require the ENABLE_COMMUNITY_SHARING setting or admin role.

Advanced Features

Message Rating & Feedback

Provide feedback on model responses:
  • Rating System: Rate messages with thumbs up/down
  • Custom Tags: Add tags to annotate message quality
  • Analytics: Track which models produce better responses

Conversation Branching

Explore different conversation paths:
  • Clone at Message: Create a new chat branching from any message
  • Original Chat Reference: Cloned chats retain reference to source
  • Branch Point Tracking: Know exactly where conversations diverged

Folder Organization

Group related chats in folders:
# API endpoint: POST /api/v1/chats/{id}/folder
{
  "folder_id": "project-alpha"
}
Folders support hierarchical organization - retrieve all chats in a folder including subfolders automatically.

Real-time Features

WebSocket Events

Stay synchronized with live updates:
  • Message Updates: Real-time message edits across sessions
  • Typing Indicators: See when models are generating responses
  • Multi-device Sync: Changes propagate instantly to all your devices

Collaborative Features

  • Shared Chats: Generate shareable links for conversations
  • Clone Shared Chats: Import interesting conversations shared by others
  • Public/Private Control: Manage sharing permissions per chat

Performance Features

Pagination

Chats are loaded efficiently with:
  • 60 chats per page in list views
  • Lazy loading for smooth scrolling
  • Skip/limit parameters for API access

Caching

Optimized for performance:
  • Chat lists cached for quick access
  • Incremental updates reduce server load
  • Smart invalidation on changes

API Reference

List Chats

GET /api/v1/chats/list?page=1
Retrieve your chat history with pagination.

Create Chat

POST /api/v1/chats/new
Start a new conversation.

Update Chat

POST /api/v1/chats/{id}
Modify chat content or metadata.

Delete Chat

DELETE /api/v1/chats/{id}
Remove a conversation permanently.
All API endpoints require authentication. See User Authentication for details.