The wait is over. Google's Nano Banana 2.1 is now live.

BytePlus OmniHuman 2.0

Image Gen·omnihuman·by Bytedance

BytePlus OmniHuman 2.0 generates animated videos from images and audio with optional motion prompts. Create character videos for presentations and social posts.

Try it now →
API reference
Runtime (p50)
-
Estimated price
$0.07 / sec
Call the API
prediction.sh
sh
curl -X POST \
  -H "Authorization: Bearer $EACHLABS_API_KEY" \
  -H "Content-Type: application/json" \
  --data '{
    "model": "byteplus-omnihuman-2-0",
    "input": {
        "prompt": "She speaks warmly and naturally to the camera with subtle head movements, expressive eyebrows and a friendly smile. The camera stays still.",
        "audio_url": "https://cdn-us.eachlabs.ai/defaults/df5d856e0ba6416482912c021aa9356b.wav",
        "image_urls": [
            "https://cdn-us.eachlabs.ai/defaults/d32d893d74314d9da5d56a033bc006b5.png"
        ]
    },
    "webhook_url": ""
}' \
  https://api.eachlabs.ai/v1/prediction/

Related models

4 models
Hotel Room Studio Dance AI model preview
Hotel Room Studio DanceUnified AI API for hundreds models. Access video, image, and audio generation from leading providers through one integration with eachlabs.
MiniMax | H3 | Max | Image to Video AI model preview
MiniMax | H3 | Max | Image to Videofal
Subway Freeze Video AI model preview
Subway Freeze VideoUnified AI API for hundreds models. Access video, image, and audio generation from leading providers through one integration with eachlabs.
Smash and Grab Video AI model preview
Smash and Grab VideoUnified AI API for hundreds models. Access video, image, and audio generation from leading providers through one integration with eachlabs.
* FAQ

About BytePlus OmniHuman 2.0

01 / 03

What does BytePlus OmniHuman 2.0 generate?

BytePlus OmniHuman 2.0 generates a video from image URLs and an audio file. You can optionally describe the scene, movement, or camera direction in a text prompt. The finished video is returned as a hosted video URL, making it easy to include in a content creation workflow.