MiniMax H3 generates video from a text prompt, from a start and/or end frame, or from a mix of reference images, videos, and audio. It runs 4 to 15 seconds at 768P or 2K across six aspect ratios (21:9, 16:9, 4:3, 1:1, 3:4, 9:16), takes prompts up to 7,000 characters, and accepts up to 9 reference images, 3 reference videos, and 3 reference audio clips in reference mode. It suits longer clips that need to stay on-model with existing footage or artwork.
MiniMax H3 generates video from a text prompt, from a start and/or end frame, or from a mix of reference images, videos, and audio. It runs 4 to 15 seconds at 768P or 2K across six aspect ratios (21:9, 16:9, 4:3, 1:1, 3:4, 9:16), takes prompts up to 7,000 characters, and accepts up to 9 reference images, 3 reference videos, and 3 reference audio clips in reference mode. It suits longer clips that need to stay on-model with existing footage or artwork.
Credit costs are examples — final credits are calculated during generation based on your selected resolution and duration.
See the full model catalog, pricing plans, or take the model quiz.
We use analytics to improve your experience. See our Privacy Policy.