Home Guides MiniMax H3 Max audio and dialogue
MiniMax H3 Max guide
MiniMax H3 Max audio and dialogue
H3 Max Studio · last updated 3 September 2026
MiniMax H3 Max predicts frames and stereo audio together. Write the spoken line, the room, and the score in the same prompt. On this site, showcase and studio clips start muted so they can autoplay; turn sound on to judge the mix.
What the audio pass can do
H3 Max can place dialogue, foley, ambience, and a short musical bed in one 5 to 15 second take. It is strongest when the prompt names the sound as clearly as the camera. Vague 'cinematic music' underperforms a specific cue.
How to prompt dialogue
Put the spoken line in quotes. Name who speaks and the room. Example: a detective under a fogged street lamp says, "The ledger was in the coat the whole time." Sound: slow footsteps on wet stone and one distant bell.
Previewing sound on this studio
Browsers block unmuted autoplay. Sample clips and generated previews therefore start silent. Use the speaker control to hear the track. Only one clip plays sound at a time.
FAQ
Does every MiniMax H3 Max video have audio?
The model generates a stereo track with the picture. Quality depends on how clearly you prompt sound.
Can I upload reference audio?
Yes in reference-to-video, together with image or video files. Audio cannot be the only reference.
Why do homepage clips start muted?
So they can autoplay. Click the speaker to unmute.