m2

The fastest local AI engine for Apple Silicon. 4.2x faster than Ollama, 0.08s cached TTFT, 100% tool calling. 17 tool parsers, prompt cache, reasoning separation, cloud routing. Drop-in OpenAI replacement. Works with Claude Code, Cursor, Aider.

reading the undocumented mems accelerometer + gyroscope on apple silicon macbooks via iokit hid

Run Qwen3-TTS text-to-speech locally on Mac (M1/M2/M3/M4). Voice cloning, voice design, custom voices. 100% offline using MLX.

A fix that adds custom resolutions, upscaling, mod support and more to Metal Gear Solid in the Master Collection, and other titles.

Run larger LLMs with longer contexts on Apple Silicon by using differentiated precision for KV cache quantization. KVSplit enables 8-bit keys & 4-bit values, reducing memory by 59% with <1% quality loss. Includes benchmarking, visualization, and one-command setup. Optimized for M1/M2/M3 Macs with Metal support.

VGGT 3D Vision Agent optimized for Apple Silicon with Metal Performance Shaders

A Home Assistant Custom Component for Maestro MCZ (cloud)

Unofficial native Mixxx builds for macOS (Apple Silicon/Intel) and Linux

China Macroeconomic Dashboard

🎨 Native AI image generation for Apple Silicon with Qwen-Image. Lightning LoRA acceleration for fast 4–8 step runs. Zero Docker, just works.

A port to Apple Silicon (see branch silicon)

The NVIDIA Container Toolkit for Mac — Give any Docker container full Apple Silicon Metal GPU access. 100+ GPU operations, LLM inference, training, image gen, audio, embeddings. Zero CUDA. Just Metal.

Python demos for testing out the Stable Diffusion's XL (SDXL 0.9) model.

Run Tencent's Hunyuan3D 2.1 image-to-3D model without CUDA — natively on Apple Silicon (MPS) and AMD GPUs (ROCm), now with PBR texturing

A python training pipeline for Apple MLX to produce a fine tuned model with Apple Neural Engine Support

Interactive demo showcasing MiniMax-M2's interleaved thinking for agentic workflows, demonstrating adaptive strategy after each tool call with transparent traces and cost benchmarking.

🍎 Accelerate 3D reconstructions from images using VGGT, optimized for Apple Silicon with Metal Performance Shaders for fast and efficient processing.

Train reasoning models on your Mac. GRPO training framework for Apple Silicon with curriculum learning.

Run CUDA-hardcoded PyTorch repos on Apple Silicon (MPS) with zero source edits

MTM2 Commandable SAL Component