vmlx

Open-source Mac app for local LLM inference on Apple Silicon

AI AgentsLLM ToolsOpen sourceOpen sourceAPI available
Visit tool

vmlx.net

About vmlx

vMLX is an open-source Mac app for running large language models (LLMs) locally on Apple Silicon. It provides features such as prefix caching, paged KV cache, continuous batching, and MCP tool integration.

Description summarised by AI from the sources listed below.

Key features

  • prefix caching
  • paged KV cache
  • continuous batching
  • MCP tool integration

Use cases

  • local LLM inference
  • LLM development
  • agentic workflows

Pricing

Pricing model: Open source. Detailed plans are not recorded; check the official website for current prices.

Pricing from the tool's own website.