MiniMax-H3 video model ported to MLX, runs on M5 Max in 45 minutes

MiniMax-H3 通过 MLX 移植可在 Apple Silicon 上运行

Simon Willison got the MiniMax-H3 MLX port running on an M5 Max MacBook Pro. The model, released two days ago, takes text, images, audio, and video as input and outputs 15-second video clips with audio. He downloaded ~115 GB of model files and generated one clip from a text prompt in just under 45 minutes. The visuals looked good, but the audio came out as garbled speech-like noise—he didn't follow the official prompting guide for audio, so that part isn't a fair test.

Why it matters: Simon Willison's first-person MLX port gives the first real local inference benchmark for MiniMax H3: 45 min on M5 Max for a 15-sec clip. Valuable data, but the model isn't a flagship release and 45-minute inference is far from practical, capping it at 78.