← Latest products

Run 500B+ Parameter LLMs Locally on a Mac Mini

By combining 1.58-bit ternary quantization, dynamic sparsity with Top-K pruning and MoE routing, and mmap-based layer streaming, OpenGraviton can run models far larger than your system RAM—even on a Mac Mini.

Open source

56 stars · GitHubexternal site

Visit github.comexternal site

Reactions (1)

Similar products

by description similarity

  1. OpenGraviton

    opengraviton.github.io

    Run 500B+ parameter models on a consumer Mac Mini

    Hacker News13 pts

  2. Ollama

    github.com

    Run LLMs on your Mac

    Hacker News284 pts

  3. 80% faster

    github.com

    50% less memory, 0% loss of accuracy Llama finetuning

    Hacker News385 pts

  4. KVSplit

    github.com

    Run 2-3x longer contexts on Apple Silicon

    Hacker News272 pts

  5. Bodhi App

    github.com

    Run Open Source/Weights HuggingFace LLMs Locally

    Hacker News21 pts