# antirez/h3.c

MiniMax H3 inference engine for Mac computers

Repository: https://github.com/antirez/h3.c
Canonical: https://ross.abutalabs.com/products/h3c
Language: C
License: MIT
License Family: permissive
Last push: 2026-08-11T13:53:32+00:00

## Health v2 (maintenance only)
Score: 56/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 97, release rhythm 35, longevity 1
- inputs: {"age_days": 24, "days_push": 22, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, young
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2443, forks 177 (observed 2026-08-28T04:06:51.423815+00:00)

## What it is
A native C inference engine for the MiniMax H3 model on Apple Silicon, using Metal for GPU acceleration. It generates video (and audio) from text prompts, with first/last-frame conditioning and reference-image (Ref2VA) support via an interactive CLI session.

## Use cases
- generate videos from text prompts on a Mac
- run MiniMax H3 locally on Apple Silicon
- animate between a first and last frame image
- generate video conditioned on reference images
- interactively iterate on video generation prompts and seeds

## When to choose
- you have a Mac with Apple Silicon (M3/M5 Max) and want local video generation
- you want a fast native Metal implementation without Python dependencies
- you need first/last-frame or reference-image conditioning for video generation

## When to avoid
- you need inference on Linux, Windows, or NVIDIA/AMD GPUs
- you want a polished GUI rather than a terminal session
- you need a production-stable, widely supported inference stack

## Facets
- artifact type: application
- maturity: active
- function: llm-inference, video-processing, image-processing, gpu-computing, cli
- domain: artificial-intelligence, machine-learning, image-processing, gpu-computing
- platform: cli, cpp
- tags: apple-silicon, metal, minimax-h3, text-to-video, diffusion-transformer, video-generation, interactive-session, video, macos, gpu

## Member repositories
- antirez/h3.c (main) score 56

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:51.423815+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:30:35.693579+00:00, confidence not recorded.
  - readme: https://github.com/antirez/h3.c (fetched 2026-08-28T04:06:51.423815+00:00, sha 310d82d398a3)
- Data as of 2026-08-30T08:39:29.467469+00:00.