DaVinci MagiHuman

davinci-magihuman

Image-to-video avatar generation with optional driving audio for lip sync. Upload audio to sync to a real recording, or omit it to generate speech from the prompt.

Open in Studio Create an API key

Pricing

Includes our margin. Charged per request from your balance.

Per second: $0.055

Supported durations: 1–30 s

Details

Provider: Gair
Added: Apr 3, 2026

Options

Parameters this model accepts.

Parameter	Values	Default
`seed` Optional seed for reproducible output.	Number	—
`duration` Length of the generated video in seconds.	1–30 second Default `5`	`5`
`audio_url` Optional audio URL for lip-sync mode. If omitted, audio is generated from the prompt.	Text	—
`resolution` Output resolution.	256p, 540p, 720p, 1080p Default `256p`	`256p`
`guidance_scale` Optional classifier-free guidance scale (0-20).	0 to 20 Default `5`	`5`
`num_inference_steps` Optional denoising steps (1-50).	1 to 50 Default `8`	`8`
`enable_safety_checker` Run prompt and image safety checks before generation.	On or off Default `on`	`on`

Pricing

Details

Options

Search