Runs AI agents entirely on your Mac — a native Apple Silicon runtime for local MLX models with shared memory, file access, and an OpenAI-compatible server. Ships CLI and MCP support. MIT, 7k+ stars.
- Access
- CLIAPIMCP
- Auth
- None
- Skills
- 3
Skills
Local Model Runtime
Runs open MLX models on Apple Silicon with no account or internet, keeping all data on the user's Mac.
OpenAI-Compatible Server
Exposes a local OpenAI-compatible API so existing clients and agent frameworks connect without code changes.
MCP Tooling
Runs as an MCP server so agents can call local tools and files through the Model Context Protocol.
Related Agents
oMLX
CLI · APIServes local LLM, vision, embedding and reranker models on Apple Silicon Macs through OpenAI- and Anthropic-compatible…
Magnitude
CLI · APIRuns open-weight models locally on kernels tuned to your hardware and connects them to coding agents through the…
TensorFold
API · CLIServes local language models from a native Zig binary on Apple Silicon and NVIDIA GPUs with OpenAI- and…
LM Studio
API · CLI · MCPRuns open LLMs locally with an OpenAI- and Anthropic-compatible server, the lms CLI, a headless daemon, and MCP…