Remy Reference/models/MiniMax H3 Max Lip Sync

MiniMax H3 Max Lip Sync

MiniMax H3 Max Lip Sync, a lipsync model from Fal.

lipsyncModelOverride: { model: "minimax-h3-max-lip-sync" }

MiniMax·served via

Fal·Released July 2026

Pricing

List price: $0.05-$0.32/second
Per second: $0.050
Per minute: $3.00

Specifications

Context window: 2,000 tokens
Aspect ratios: 4
Released: July 2026
Model string: minimax/h3-max/lip-sync/image-to-video

Parameters

Model options, passed in lipsyncModelOverride.config.

Parameter	Type	Default	Range / options
`image` Portrait to animate into a talking video, with an aspect ratio between 0.4 and 2.5.	Image URL	—	—
`audio` Speech to lip sync to, from 5 seconds to 15 minutes. The output runs as long as the audio.	Audio URL	—	—
`resolution` The resolution of the output video.	Select	1080P	480P768P1080P2K
`enable_transcription` Transcribe the speech first and use the transcript to guide the lips.	Boolean	false
`seed`	Seed	—	—

Use this model

await mindstudio.generateLipsync({
  lipsyncModelOverride: {
    model: "minimax-h3-max-lip-sync",
    config: {
      image: "...",
      audio: "...",
      resolution: "1080P",
    },
  },
});