ByteDance OmniHuman 2.0: AI portrait animation
One photo. One audio track. A 720p video of that person performing your audio — face, head, and body moving to the sound, not just the lips. ByteDance OmniHuman 2.0 goes beyond lip-sync to animate the whole performance, up to 5 minutes long in a single continuous shot. Available now in Picsart's AI Playground.
Performances made with ByteDance OmniHuman 2.0
What is ByteDance OmniHuman 2.0?
ByteDance OmniHuman 2.0 is an AI portrait animation model that turns one photo and one audio track into a 720p video of that person performing the audio. It goes beyond lip-sync — the face, head, and body all move to the sound — so a still image becomes a full performance, not just a moving mouth. The input photo can be a portrait, selfie, character, or illustration, and the audio can be speech, dialogue, or music. Clips run from 4 seconds up to 5 minutes in a single continuous shot, with no stitching. It's the next generation of ByteDance's OmniHuman human-animation family, available in Picsart's AI Playground.
- One photo + one audio track
- 720p, up to 5 minutes in one shot
- Keep editing in Picsart
Beyond lip-sync: the whole performance
OmniHuman 2.0 doesn't just move the mouth. Give it a photo and an audio track and the face, head, and body all respond to the sound, so the character genuinely performs your audio instead of mouthing along. Because the output runs as one continuous shot from 4 seconds up to 5 minutes — where the previous generation maxed out around 60 seconds — a full song, speech, or podcast segment plays through in a single take, no stitching clips together. Swap in a different track and the same photo delivers a new performance.
It reads tone, emotion, and words
OmniHuman 2.0 listens to how your audio sounds, not just what it says. A whisper, a shout, and a held note each produce a different performance from the same photo — the model matches head motion, gestures, and expression to the emotion and delivery in the track. That makes a calm narration feel measured and an energetic song feel lively, all from one still image. At $0.07 per second, you can turn a single photo into a presenter, a performer, or a character that truly acts.
ByteDance OmniHuman 2.0 in Picsart AI Playground
ByteDance OmniHuman 2.0 lives in Picsart's AI Playground alongside many other AI models, so you can animate a photo and keep refining it without juggling separate tools — no third-party API keys or separate subscriptions. Reach it whichever way you work: on the web, in the desktop app, and built into your own projects via CLI, MCP, REST API, and SDK, so a performance you prototype here drops straight into an automated pipeline.
How to animate a portrait with ByteDance OmniHuman 2.0
Upload a photo, add your audio, and generate a 720p performance up to 5 minutes long — all in Picsart.
How to
Upload a photo
Bring in one image of the person or character you want to animate — a portrait, selfie, character, or illustration works. Tip: use a clear, well-lit photo with the face visible.
ByteDance OmniHuman 2.0 FAQ
Answers about AI portrait animation in Picsart.
ByteDance OmniHuman 2.0 is an AI portrait animation model that turns one photo and one audio track into a 720p video of that person performing the audio. The face, head, and body all move to the sound, and clips run from 4 seconds up to 5 minutes in a single continuous shot. It's available in Picsart's AI Playground.
More AI video models to use
Explore other AI video models available in Picsart AI Playground.
Pricing
ByteDance OmniHuman 2.0 costs $0.07 per second of generated video — create a one-off performance or build a repeatable portrait-animation workflow for launches, social, and paid campaigns.
Pro
Most popularAI tools for everyday creative work.
Ultra
Most powerfulHeavy AI usage for creators & teams.
Read up on AI video
Guides and tutorials on making and editing AI video in Picsart.
