Livepeer Agent · Fal Expansion · video · native audio

Multimodal Reference Trailer · MiniMax H3

Combine up to 9 images, 3 video clips, and 3 audio clips as references — MiniMax H3 generates video with native stereo audio and full multi-reference continuity.

Format
Video · native audio
Caps
minimax-h3-ref2v · h3-t2v · h3-i2v
Cost
~$0.16/s (768p)
Persona
Director · Brand studio

MiniMax H3 ref2v output — multi-reference video generation with native stereo audio. Subjects, motion, and audio stay consistent across references.

Why this works: MiniMax H3 is the first open-weight multimodal video model with native stereo audio on every generation. Combine up to 9 images, 3 video clips, and 3 audio clips — the model holds subjects, motion, and audio consistent without separate compositing passes.
Open playbook → ← All playbooks All showcases →