OmniHuman 2 Pro
MOST POPULAR$19.99/ week
Billed weekly. Cancel anytime.
1,200 credits every week≈ 120 seconds of video
- Full OmniHuman 2.0 model, 720p
- Videos up to 5 minutes each
- Prompt control for gestures & camera
- No watermark · commercial use
NEWOmniHuman 2.0 model — videos up to 5 minutes, multi-person scenes
OmniHuman 2.0One photo + any voice = a video that talks, sings and moves.
Lifelike lip sync, natural gestures, camera motion and multi-person performances — generated from a single image and an audio clip with the OmniHuman 2 model.
STEP 2 · VOICEAdd a voice or songMP3 or WAV · 4 s to 5 min
- Free to upload & preview
- Sign in with Google
- Your uploads stay private
- Max video length per render
- 5 min
- Native output resolution
- 720p
- Prompt languages supported
- 6
- Is all it takes — plus audio
- 1 photo
Showcase
Made with OmniHuman 2.0
Every clip below started as one still image and one audio file. Pick a style and drop in your own photo.
How it works
How to use OmniHuman 2.0 in three steps
- 01
Upload one photo
A portrait, a full-body shot, a group photo, even a 3D character or a pet. Faces should be clear and unobstructed.
- 02
Add a voice or a song
Upload an MP3 or WAV, or record straight from your browser. Anything from 4 seconds up to 5 minutes.
- 03
Direct it, then generate
Optionally describe gestures, mood or camera moves. OmniHuman 2.0 renders a 720p video synced to every syllable.
OmniHuman 2.0 vs 1.5
What changed in 2.0
Same input, two generations. 2.0 handles longer takes, several people in one frame and prompt-driven camera work.
| Capability | OmniHuman 1.5 | OmniHuman 2.0 |
|---|---|---|
| Max audio / video length | 30 s (1080p) · 60 s (720p) | Up to 300 s |
| People in one frame | One speaker (mask selects who) | Multi-person performances |
| Camera, body & gestures | Prompt guidance | Camera, body & gesture control |
| Output | 720p or 1080p | 720p native |
Use cases
One model, a whole cast
UGC video adsTest hooks with creator-style talking heads without booking a shoot.
AI spokespersonProduct explainers, onboarding and announcements in your brand’s voice.
Singing avatarsMusic videos and covers with breath, rhythm and emotion in the performance.
Podcasts & dialoguesTwo or more people in one frame, each speaking their own lines.
Talking photosBring a portrait, a painting or an old family photo to life.
Characters & mascots3D, anime and animal characters that gesture and emote on cue.
Pricing
OmniHuman 2.0 pricing
10 credits = 1 second of video. Upload and preview free — pay only when you render.
OmniHuman 2 Ultra
BEST PER SECOND$49.99/ week
Billed weekly. Cancel anytime.
3,500 credits every week≈ 350 seconds of video
- Everything in Pro
- Lowest price per second
- Priority rendering queue
- Built for agencies & heavy creators
Subscriptions renew automatically until cancelled. Cancel anytime from your account. Unused credits reset each period.
Guides
All guides →Learn OmniHuman 2
OmniHuman 2.0 FAQ
What is OmniHuman 2.0?
OmniHuman 2.0 is ByteDance’s latest audio-driven human video model. It turns one image and one audio clip into a realistic video with lip sync, facial expression, gestures, body and camera movement.
How is OmniHuman 2.0 different from OmniHuman 1.5?
Version 2.0 accepts audio up to 5 minutes, supports multi-person performances in one frame and gives you prompt control over camera, body movement and gestures. OmniHuman 1.5 was limited to 30 seconds at 1080p or 60 seconds at 720p.
Is OmniHuman 2 free?
Uploading your photo and audio and previewing your project is free. Rendering the final video uses credits from a Pro or Ultra plan — 10 credits per second of video.
What photos work best?
A sharp, well-lit image where faces are clearly visible. Portraits, half-body and full-body shots, group photos, illustrations and 3D characters all work.
Can I use the videos commercially?
Yes. Videos rendered on a paid plan have no watermark and can be used in ads and client work, provided you own the rights to the photo and audio you upload.
Is this the official OmniHuman website?
No. omnihuman2.ai is an independent service that runs the OmniHuman 2 model through a licensed API provider. We are not affiliated with ByteDance or BytePlus.
Your photo is one click from talking.
Upload now — it’s free to preview.