Agent: Cursor, Claude CodeLLM: Qwen3, DeepSeek#local-llm#apple-silicon#mlx#tool-calling#openai-compatible
Rapid-MLX lets you run LLMs locally on your Mac with blazing speed — 0.08s cached TTFT, 100% tool calling support, and 17 tool parsers. It's a drop-in OpenAI API replacement that works natively with Cursor, Claude Code, Aider, LangChain, and PydanticAI.