# H3 Max Lip Sync

Give it one photo of a person and a recording, and it makes them say it. Up to 2K, and the clip is as long as the sound.

Model id: `h3-max-lipsync`
Type: Video
Made by: MiniMax
Inputs: First frame, Audio
Length: 5 to 15 s
Resolution: 480P, 768P, 1080P, 2K
Updated: Oct 12, 2026

## Pricing

| Setting | MiniAPI | fal | Difference |
| --- | ---: | ---: | ---: |
| 480p | $0.055 / s | $0.05 | +10% |
| 768p | $0.088 / s | $0.08 | +10% |
| 1080p | $0.176 / s | $0.16 | +10% |
| 2K | $0.352 / s | $0.32 | +10% |

You are not charged for a job that fails.

## Parameters

| Name | Required | Values | Default | Description |
| --- | --- | --- | --- | --- |
| `image_url` | Yes | https URL |  | A picture to start from. We fetch it and keep a copy for the job. |
| `audio_url` | Yes | https URL |  | The sound to use. |
| `resolution` | No | 480P, 768P, 1080P, 2K | 768P | Output size. Higher is sharper, slower and costs more. |
| `enable_transcription` | No | true, false | false | Transcribe the supplied audio to guide lip synchronization. When disabled, synchronize to the audio without a transcript. |
| `seed` | No | 0 to 2147483647 |  | The same seed with the same settings gives a similar result. Leave it out for a random one. |

Page: https://api.minicut.ai/models/h3-max-lipsync
API reference: https://api.minicut.ai/docs/api