BigHugger
GH Repository · AtomicBot-ai

Atomic-Chat

Local AI app and inference engine for agents. Run open-weight LLMs locally — private, 100% offline on your computer. Join our Discord: https://discord.com/invite/8wGSsvmg4V

stars
1,485
30-day movement
+196/day
Related entries
60
Connections
3
makemcp-serverqwenopen-sourceself-hostedmlxnode/yarnnodemcplocal-llmlocal-firstai-chatapple-silicondeepseekchatgptlocal-aigemmaggufTypeScriptllamacppgpt-ossai-toolsllmdesktop-app

Atomic-Chat is a local AI desktop app and inference engine for running open-weight LLMs entirely offline on your own computer. It is TypeScript-based, exposes an MCP server role, and supports GGUF-based runtimes like llama.cpp and MLX on Apple Silicon.

You want private, fully offline local LLM inference with agent/MCP support in a desktop app.

Use it to

  • Run open-weight models like Qwen, Gemma, or DeepSeek locally
  • Serve an MCP server for agent tooling
  • Chat with local LLMs without any network access
  • Use Apple Silicon acceleration via MLX or llama.cpp
  • Self-host a local-first AI chat desktop app

For Developers and users wanting offline local LLMs

Role
mcp-server
Language
TypeScript
Licence
custom (licence file present)
Forks
176
Open issues
44
Last push
2026-09-16
Latest release
v1.1.17 · 2026-03-31
topicslocal-llmmcp-serverllm-inferencedesktop-appapple-siliconself-hosted