PAS

Personal AI Workstation — a lightweight web app for chatting with multiple AI providers from a single interface.

Features

• Multi-platform support — OpenAI-compatible APIs, Anthropic Claude, Ollama, and local models via llama.cpp
• Local model inference — Run Qwen2.5-3B on-device with llama.cpp, auto-downloaded from ModelScope
• Streaming responses — Real-time token-by-token output
• Configurable providers — Add, switch, and manage AI services from the sidebar
• Customizable settings — Temperature, max tokens, and system prompt per session

Quick Start

uv run uvicorn app.main:app --host 0.0.0.0 --port 8000 --reload

Then open http://localhost:8000 in your browser.