Cost
$2.10/minute · $0.75/minute ($0.0125/sec)
MiniMax FastH3 Omni
NEW$2.10/MINUTE$0.75/MINUTE
Faster-than-realtime video generation with text, audio and image input.
FastH3 Omni makes video with sound. Start from a prompt alone, or add up to 9 pictures of people, objects, or places to put them in the video. With at least one picture, you can add up to 3 audio clips. Each clip is a person's voice, the soundtrack a face lip-syncs to, or background sound.
The model makes the sound together with the video, so one prompt describes both. Up to 3 pictures keep the clip in real time. With 4 to 9, the clip takes longer to build than to play.
This FastH3 Omni site is an independent product demonstration. The API tab mirrors the public model workflow and does not issue live inference credentials.
Speed
Faster than realtime
Pictures
Up to 9 of people, objects, or places
Audio Clips
Up to 3, each a voice, soundtrack, or background sound
Video Length
5.875, 8, 10.125, 12.25, or 14.375 seconds
In
Text Prompt, Pictures, Audio Clips
Out
Generated Video, Generated Audio
Transport
Streaming tracks for video and audio