Open-source tools to run AI on your own machine

Your data stays local. No subscription.

1. rtkrtk-ai/rtkNew★ 84.5kApache-2.0Rust

   CLI proxy that reduces LLM token consumption by 60-90%

   Single Rust binary, zero dependencies

   brew install rtk # macOS / Linux winget install rtk-ai.rtk # Windows curl -fsSL https://raw.githubusercontent.com/rtk-ai/rtk/refs/heads/master/install.sh | shCopy
2. unslothunslothai/unslothNew★ 78.5kApache-2.0Python

   GUI app to run and train large language models and diffusion models locally

   Supports multiple model formats and GPU acceleration

   curl -fsSL https://unsloth.ai/install.sh | shCopy
3. Ollamaollama/ollama★ 185kMITGo

   Download and run open-weight models like Qwen, DeepSeek, Gemma and gpt-oss locally with one command, with an OpenAI-compatible API.

   The backend behind almost every local AI tool: once installed, browser extensions, note plugins and coding agents can connect to it directly.

   brew install ollamaCopyLinux：curl -fsSL https://ollama.com/install.sh | sh
4. Open WebUIopen-webui/open-webui★ 157kCustom licensePython

   A self-hosted ChatGPT-style web UI that connects to Ollama and any OpenAI-compatible API, with built-in RAG, web search, multi-user support and MCP.

   Good for running on a home server or NAS for a family or small team; the most complete of the self-hosted UIs.

   pip install open-webui && open-webui serveCopyor Docker
5. Cherry StudioCherryHQ/cherry-studio★ 52.9kAGPL-3.0TypeScript

   A cross-platform desktop AI client that puts many cloud models and local Ollama in one interface, with a built-in knowledge base, MCP, agents and 300+ assistant templates.

   Built by a Chinese team with complete Traditional and Simplified Chinese UIs. Supports Chinese models like DeepSeek, Qwen and Kimi as well as international ones; easy to set up.

   InstallDownload the installer from cherryai.com or GitHub Releases
6. Janjanhq/jan★ 45.1kCustom licenseRust

   A ChatGPT-style desktop app that runs fully offline, with built-in model downloads and the llama.cpp engine; can also switch to cloud APIs.

   The friendliest local LLM option for people who'd rather avoid the terminal. Small Tauri build, and it doubles as a local API server.

   brew install --cask janCopy
7. Handycjpais/Handy★ 34.5kMITRust

   Press a hotkey and speak; local Whisper/Parakeet models transcribe your speech and type it into any app.

   Fully offline and free — an open-source alternative to paid dictation tools like Superwhisper and Wispr Flow. Supports Chinese.

   InstallDownload the macOS/Windows/Linux installer from GitHub Releases

Updated Oct 25, 2026 · ★ GitHub