Skip to content

OmniHuman 2.0 AI video generator

Lip synced avatar video from a portrait. Make 4 to 300s clips up to 720p with native audio in your browser, and see the credit price before you generate.

Model

New and popular

All models

Lip synced avatar video from a portrait

Reference images0/2 images
Settings5s · 720p · auto · Native audio
Settings
Resolution
Aspect ratio

Credits come back automatically if a job fails.

Example

Reference to video example
Try a prompt like The woman from the reference walks through a night market, handheld camera

What you can do with OmniHuman 2.0

What this model can do here, taken from its live settings. Try any of them in the generator above.

  • Up to 2 reference files

    Add several images of the same character or product to keep them consistent from shot to shot.

  • Native audio in the same pass

    Turn on audio and the clip comes back with sound that matches the action, with no separate sound step.

  • Pick length and resolution

    Clips run 4 to 300s. Output from 720p to 720p.

How to use OmniHuman 2.0

  1. 01

    Pick an input

    Start from reference to video.

  2. 02

    Describe the result and set options

    Write what you want to see, then choose length, resolution. The credit price updates as you go.

  3. 03

    Generate and download

    Most clips finish in one to three minutes. Download the file, or tweak the prompt and run it again.

OmniHuman 2.0 pricing

Longer clips and higher resolution cost more. The generator shows the exact price for your settings before anything is charged.

Starting price
Currently unavailable
In dollars
—
Free to try
48 credits on sign up, no card
See credit packs and plans

OmniHuman 2.0 compared with similar models

Live specs and starting prices. Click a model to open its page.

OmniHuman 2.0 compared with Kling 4.0, Seedance 2.5, Veo 3.1
SpecOmniHuman 2.0Kling 4.0Seedance 2.5Veo 3.1
MakerByteDanceKuaishouByteDanceGoogle
Starting price—194 credits116 credits192 credits
InputsReferenceText · First and last frame · ReferenceText · First and last frame · ReferenceText · First and last frame
Max input files215502
Duration4 to 300s3 to 30s4 to 30s4 to 8s
Resolution720p720p · 1080p480p · 720p · 1080p720p · 1080p · 4K
Aspect ratioauto16:9 · 9:16 · 1:1 · 21:9 · autoauto · 21:9 · 16:9 · 4:3 · 1:1 · 3:4 · 9:1616:9 · 9:16
Native audioYesYesYesYes

Full OmniHuman 2.0 settings

Show

Reference

Up to 2 files

Duration
4 to 300s
Resolution
720p
Aspect ratio
auto
Native audio
On or off

OmniHuman 2.0 questions

What is OmniHuman 2.0?

OmniHuman 2.0 is ByteDance's AI video model: lip synced avatar video from a portrait. On FrameMux you run it in the browser with credits, next to every other model.

Is OmniHuman 2.0 free to try?

Yes. New accounts get 48 free credits with no card.

How much does OmniHuman 2.0 cost?

Longer clips and higher resolution cost more, and the exact price shows before you generate. Credit packs start at $9.90 for 990 credits.

What inputs does OmniHuman 2.0 accept?

Reference. You can upload up to 2 files per job. JPG, PNG and WebP images up to 10 MB each. The panel only shows the options this model accepts.

Which lengths, resolutions and aspect ratios does OmniHuman 2.0 support?

Length: 4 to 300s. Resolution: 720p. Longer and sharper outputs cost more credits.

Does OmniHuman 2.0 generate audio?

Yes. Turn on native audio in the settings and the clip comes back with matching sound. Leave it off to save credits when you plan to add your own track.

How do I keep a character or product consistent with OmniHuman 2.0?

Use reference mode and upload up to 2 clear images of the same subject from different angles. Describe the subject the same way in every prompt.

Do I need a ByteDance account or API key?

No. You use OmniHuman 2.0 in the browser with FrameMux credits. There is no API key to set up and no separate subscription.

Can I use OmniHuman 2.0 results commercially?

Paid credit packs and plans include commercial use, within ByteDance's terms for OmniHuman 2.0.

Can I upload photos of real people?

Only photos you have permission to use. Content of real people without consent and adult content are blocked, and blocked prompts never use credits.

Are my OmniHuman 2.0 results private?

Yes. Uploads and results are private to your account and are deleted after 30 days, so download what you want to keep.

What happens if a OmniHuman 2.0 job fails?

The credits return to your balance automatically, and the error tells you what to change.

Start your first OmniHuman 2.0 clip

48 free credits, price shown first, credits back if a job fails.

Create with OmniHuman 2.0