A contemporary web interface for interacting with Ollama models, built with Flask and vanilla JavaScript. Supports multimodal conversations with file attachments, theme customization, and persistent conversation history.

I recently noticed a look-alike of this repository that lists me as a contributor (probably by someone pulling this and using the commit history info). The repository is found here: github.com/TzerAWK/woolychat_releases

They have a directory called weightometer with a zip file called woolychat_releases.zip in the repository. The zip file contains several executables. It appears to include a pre-compiled LuaJIT standalone interpertor for Windows and has a script intended to call an executable passing the name of a text file as an argument. The text file appears to contain Lua bytecode. So basically just beware of that repo and don't download anything from it. I have reported it to Github.

I am planning to add things like Projects, Project Knowledge, folders, and unstructured tagging, but who knows when that'll happen. I only went down this rabbit hole because Open WebUI stopped letting me login locally and I didn't get why I needed an account with them in the first place.

Available as Mac and Windows apps: Releases

• Real-time Streaming Chat - Live responses from Ollama models with typing indicators
• Multimodal Support - Send images, documents, and text files to vision-capable models
• File Upload System - Drag & drop or click to upload with 5MB limit
• Conversation Management - Persistent chat history with SQLite database
• Multiple Model Support - Switch between any locally available Ollama models

• Supported Formats:
  • Images: JPG, PNG, GIF, BMP, WebP
  • Documents: PDF, DOC, DOCX, TXT, MD, RTF
  • Data: CSV, XLS, XLSX, JSON, XML, HTML
• Smart Processing:
  • Images sent as base64 to vision models
  • Text extraction from documents
  • Content preview and metadata display
• Visual Feedback: Real-time upload progress and status

• 9 Beautiful Themes - Ocean Breeze, Golden, Autumn, Midnight, Cyberpunk, Vinyl, Forest, Moonlit Lilac, Zebra, Koala
• Responsive Design - Mobile-friendly interface with adaptive layouts
• Settings Panel - Theme selection, markdown toggle, auto-scroll options
• Drag & Drop Interface - Intuitive file attachment workflow

woolychat/
├── ollama_chat.py          # Main Flask application
├── models.py               # SQLAlchemy database models
├── utils/                  # Modular utility classes
│   ├── __init__.py
│   ├── file_manager.py     # File upload/validation/storage
│   ├── text_extractor.py   # Content extraction from files
│   └── conversation_manager.py # Database operations
└── uploads/                # File storage directory

static/
├── chat.js                 # Main application logic
├── style.css               # Comprehensive styling
└── img/
    └── woolychat.png       # Application logo

-- Users table
User: id, username, email, created_at

-- Conversations table  
Conversation: id, user_id, title, model_name, created_at, updated_at, 
              message_count, is_archived, is_favorite

-- Messages table
Message: id, conversation_id, role, content, created_at, message_metadata

-- File attachments table
MessageAttachment: id, message_id, filename, original_filename, 
                   file_path, file_size, mime_type, uploaded_at,
                   extracted_text, is_processed, processing_error

-- Tags and Projects (for future organization features)
Tag: id, name, color, created_at
Project: id, user_id, name, description, color, created_at

1. Client-side validation - File type, size, and format checking
2. Multipart upload - FormData sent to /api/upload endpoint
3. Server processing - File storage, text extraction, metadata creation
4. Database persistence - Attachment linked to message when sent
5. Ollama integration - Images as base64, text content in message

# 1. File Analysis
if mime_type.startswith('image/'):
    # Convert to base64 for vision models
    images_base64.append(base64_data)
else:
    # Extract text content for context
    text_content += extract_text_from_file(file_path, mime_type)

# 2. Ollama API Request
ollama_request = {
    'model': model_name,
    'messages': conversation_history + [{
        'role': 'user',
        'content': text_content,
        'images': images_base64  # Only for image attachments
    }],
    'stream': True
}

• CSS Custom Properties - Dynamic theme switching via root variables
• Context Processor - Server-side theme injection into templates
• Session Storage - Persistent theme preferences
• Live Preview - Real-time theme switching in settings

woolychat/
├── ollama_chat.py              # Main Flask app (routing, Ollama proxy)
├── models.py                   # Database models and relationships
├── requirements.txt            # Python dependencies
├── utils/
│   ├── __init__.py            # Utility module exports
│   ├── file_manager.py        # FileManager class
│   ├── text_extractor.py      # TextExtractor class
│   └── conversation_manager.py # ConversationManager class
├── templates/
│   ├── base.html              # Base template with theme support
│   └── index.html             # Main chat interface
├── static/
│   ├── style.css              # Complete styling system
│   ├── chat.js                # Frontend application logic
│   └── img/
│       └── woolychat.png      # Application logo
├── uploads/                   # File upload directory (auto-created)
└── ollama_chat.db            # SQLite database (auto-created)

• GET / - Main chat interface
• GET /api/tags - List available Ollama models
• POST /api/chat - Streaming chat with file attachment support
• GET /health - Application health check

• POST /api/upload - File upload endpoint
• GET /api/files/<filename> - Serve uploaded files

• GET /api/conversations - List user conversations
• POST /api/conversations - Create new conversation
• GET /api/conversations/<id> - Get conversation with messages
• PUT /api/conversations/<id> - Update conversation metadata
• DELETE /api/conversations/<id> - Delete conversation

• GET/POST /api/settings/theme - Theme preferences
• GET/POST /api/tags - Conversation tags
• GET/POST /api/projects - Project management

THEMES = {
    'zebra': {
        'header_gradient_left': '#212121',
        'header_gradient_right': '#24293b', 
        'bg_gradient_left': '#1e2936',
        'bg_gradient_right': '#28303f'
    },
    'koala': {
        'header_gradient_left': '#8d9db6',
        'header_gradient_right': '#a8b5c8',
        'bg_gradient_left': '#7d8471', 
        'bg_gradient_right': '#9db4a0'
    }
    # ... 7 more themes
}

• File Type Validation - Whitelist-based MIME type checking
• File Size Limits - 5MB upload restriction
• Secure Filenames - UUID-based storage to prevent conflicts
• Input Sanitization - HTML escaping for user messages

• Sidebar Navigation - Fixed 300px conversation list
• Centered Chat - Optimal reading width with padding
• Enhanced Input - Larger attachment buttons and input area

• Collapsible Sidebar - Overlay navigation
• Full-Width Chat - Maximum screen utilization
• Touch-Friendly - Larger buttons and touch targets

# Clone repository
git clone https://github.com/haasr/woolychat.git
cd woolychat

# Create virtual environment
python -m venv .venv       # On Mac/some Linux: python3
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt

# Start Ollama (separate terminal)
ollama serve

# Run application
python ollama_chat.py

• Database: SQLite (auto-created on first run)
• Ollama URL: http://localhost:11434 (configurable in ollama_chat.py)
• Debug Mode: Enabled by default for development

1. Install Ollama - Ensure Ollama is running on localhost:11434
2. Clone Repository - Get the WoolyChat codebase
3. Install Dependencies - pip install -r requirements.txt
4. Run Application - python ollama_chat.py
5. Access Interface - Navigate to http://127.0.0.1:5000 or whichever port was selected

1. Select an Ollama model from the dropdown
2. Type message and press Enter or click Send
3. View streaming response with markdown formatting

1. Click paperclip icon or drag files onto chat area
2. Files appear in attachment preview with status
3. Send message - files are analyzed and included in conversation
4. Images are sent to vision models, documents provide text context

1. Click settings gear icon
2. Select theme from dropdown
3. Preview changes in real-time
4. Save to persist across sessions

WoolyChat - It's not Open WebUI, but it's not Open WebUI 🐑✨