H3-metal – Native MiniMax-H3 inference for Apple Silicon
Why CEREBRO kept it
LLM inference optimization for Apple
The text below is an automated extraction of the article at https://github.com/antirez/h3.c, stored verbatim in the public cerebro-vault repository. Copyright remains with the original publisher (github.com).
Native MiniMax-H3 inference for Apple Silicon. The project is being built as a sequence of working vertical slices: deterministic host/model metadata first, then portable Metal block parity, prompt encoding, prompt-to-video/audio, and first/last-frame conditioning and then ordered references. Prompt-to-video/audio, first/last-frame conditioning, and ordered Ref2VA image/video/audio references work end to end. The current work is incremental H3-specific Metal performance and memory optimization on M3 Max and M5 Max. The examples assume that the Hugging Face snapshot is in ./MiniMax-H3 and that
Community take
MiniMax H3 runs on 64GB Apple Silicon via quantization, but requires modifying workflows like ComfyUI.
Backlinks
Appeared in 1 briefing