Ollama Web Interface

A modern web interface for managing and interacting with Ollama models on localhost.

Features

🎨 Modern Dark Theme UI

• Sleek glassmorphism design with gradient accents
• Fully responsive - optimized for desktop, tablet, and mobile
• Touch-friendly controls with proper hit targets

🔄 Model Management

• Monitor Running Models - View all running models with resource usage
• Start/Kill Ollama - Control the Ollama service directly from the UI

⬇️ Smart Downloads

• Concurrent Downloads - Download multiple models simultaneously
• Real-Time Progress - Live progress bars with speed and size info (e.g., "943 MB/8.6 GB | 3.2 MB/s")
• Cancel Anytime - Stop downloads mid-process
• Persistent Progress - Downloads continue in background, visible to all users
• Auto-Cleanup - Completed/failed downloads automatically removed

🛠️ Per-Model Actions

• Info - View detailed model information in modal
• Copy - Duplicate models with new names
• Remove - Delete models with confirmation

Installation

Via Pinokio (Recommended)

1. Clone or place this repository in your Pinokio API directory: C:\pinokio\api\Ollama_Z.git\
2. Open Pinokio and find "Ollama Web Interface" in your apps
3. Click "Install" to set up the environment
4. Click "Start" to launch the web interface

Manual Installation

1. Ensure Python 3.8+ is installed

2. Create a virtual environment:

   python -m venv venv

3. Activate the virtual environment:

   • Windows: venv\Scripts\activate
   • Linux/Mac: source venv/bin/activate

4. Install dependencies:

   pip install -r requirements.txt

Usage

Start the Server

Via Pinokio: Click the "Start" button in Pinokio

Manually:

python app.py

The web interface will be available at: http://localhost:11435

Prerequisites

• Ollama must be installed and running on localhost:11434
• Download Ollama from: https://ollama.ai

Available Ollama Commands

The interface supports all major Ollama operations:

• serve - Start ollama (if not already running)
• pull - Download models from registry with progress tracking
• list - View all installed models
• ps - Monitor running models with resource info
• run - Chat with models through the interface
• stop - Stop running models
• cp - Copy models
• rm - Remove models
• show - Display detailed model information

Project Structure

Ollama_Z.git/
├── app.py              # Flask application
├── templates/
│   └── index.html      # Web interface
├── requirements.txt    # Python dependencies
├── pinokio.js         # Pinokio configuration
├── install.json       # Pinokio install script
├── start.js         # Pinokio start script

API Endpoints

Model Management

• GET /api/ps - List running models with resource info
• GET /api/list - List all installed models
• POST /api/stop - Stop a specific running model
• POST /api/rm - Remove a model
• POST /api/cp - Copy a model to new name
• POST /api/show - Get detailed model information

Download Operations

• POST /api/pull - Start downloading a model
• GET /api/pull/all - Get all active download progress (real-time)
• GET /api/pull/progress/<model> - Get specific model download progress
• POST /api/pull/cancel - Cancel an active download

Chat Operations

• POST /api/chat - Stream chat responses (Server-Sent Events)
• POST /api/chat/stop - Stop active generation

Ollama Control

• POST /api/start - Start Ollama service
• POST /api/kill - Kill all Ollama processes
• POST /api/serve - Start Ollama serve (deprecated, use /api/start)

Technical Details

Architecture

• Backend: Flask with threading for concurrent operations
• Frontend: Vanilla JavaScript (no build step required)
• Streaming: Server-Sent Events (SSE) for chat responses
• Progress Tracking: Global shared state with auto-cleanup
• Process Management: Direct subprocess control for Ollama CLI

Configuration

The server runs on 0.0.0.0:11435 by default and connects to Ollama at localhost:11434.

To change these settings, edit app.py:

OLLAMA_URL = "http://localhost:11434"
app.run(host='0.0.0.0', port=11435)

Browser Compatibility

• Modern browsers with ES6+ support
• ReadableStream API for streaming responses
• CSS Grid and Flexbox for responsive layout

License

MIT

Credits

Built for use with Ollama and Pinokio