OpenAppsSubmit

MTPLX

by youssofal

AI & LLMmacOSDesktop

The fastest way to run Qwen 3.8 Flash Next, Qwen 3.8 27B and Ternary Bonsai 2 27B on a Mac: 125 tok/s in OpenCode on an M5 Max, and a 27B model on 16 GB Macs.

The fastest way to run Qwen 3.8 Flash Next, Qwen 3.8 27B and Ternary Bonsai 2 27B on a Mac: 125 tok/s in OpenCode on an M5 Max, and a 27B model on 16 GB Macs. Native MTP speculative decoding on Apple Silicon, exact at any temperature. OpenAI and Anthropic compatible local server.

Visit websiteView on GitHubDownload
MTPLX on GitHub

Tags

#anthropic-compatible#apple-silicon#claude-code#flash-next#inference-engine#llm-inference#local-ai#local-llm

Similar apps

More in this category
  • SlotstreamAI & LLM

    Run a 105 GB AI model on a Mac that can't hold it. Slotstream streams Qwen3.8-Flash-Next (125B mixture of experts) from your SSD and caches the busiest experts…

    453SwiftMITmacOSDesktop
    1mo ago
  • Mlx DsparkAI & LLM

    Up to 4× faster LLM decoding on Apple Silicon, lossless. Native MLX port of DeepSeek's DSpark & z-lab's DFlash speculative decoding — Gemma-4, Qwen3.8,…

    702PythonMITmacOSDesktop
    3mo ago
  • Turbo FieldfareAI & LLM

    Gemma 4 26B-A4B inference in ~2 GB of RAM on any M-series MacBook

    6.9kSwiftApache-2.0macOSDesktop
    3mo ago
  • Rapid MLXAI & LLMalt. to Lm Studio

    Rapid-MLX is an open-source (Apache 2.0) OpenAI- and Anthropic-compatible LLM inference server for Apple Silicon, built on MLX, focused on reliable tool…

    4kPythonmacOSDesktop
    8mo ago
  • Vllm MlxAI & LLM

    High-performance OpenAI and Anthropic compatible LLM inference server for Apple Silicon.

    1.6kPythonApache-2.0macOSDesktop
    10mo ago
  • BackburnerAI & LLM

    Your iPhone helps your Mac run a 27B model: faster prompt reading and more context over a USB-C cable

    920PythonMITmacOSDesktopiOS
    8d ago