OmniHuman 2.0 Audio to Video API

bytedance/omnihuman-2.0/audio-to-video

720p · 4–300 sec audio

OmniHuman 2.0 Audio to Video turns one character image and 4–300 seconds of speech or music into a 720p video, with lips, expression and gestures following the audio. The image keeps the character's look and framing, while an optional prompt adds camera moves and actions.

Input

Image Urls

Drag and drop files, paste from clipboard, or provide a URL. Required. One JPEG, PNG or WebP character image.

1/1 files added

Audio Url

Drag and drop files, paste from clipboard, or provide a URL. Required. One WAV or MP3 file, 4–300 seconds, up to 15 MB.

1/1 files added

Prompt

OutputExample

Continue with

Kling Avatar 2.0 Pro Seedance 2.5 Image to Video Vidu Q4 Reference to Video

REST API

Quick Start

Make your first OmniHuman 2.0 request in three steps; the example sends an image, a voice track and a short camera prompt.

Connect to the Vidgo API

Create an API key, keep it on your server, and send it in the Authorization header as a Bearer token.

Endpoint: POSThttps://api.vidgo.ai/api/generate/submit
Authentication: Authorization: Bearer VIDGO_API_KEY

Submit one generation task

Run the short example with one image and one driving audio file. The prompt is optional.

Submit request

curl --request POST \
  --url "https://api.vidgo.ai/api/generate/submit" \
  --header "Authorization: Bearer $VIDGO_API_KEY" \
  --header "Content-Type: application/json" \
  --data '{
  "model": "bytedance/omnihuman-2.0/audio-to-video",
  "input": {
    "image_urls": [
      "https://cdn.vidgo.ai/apis/models/bytedance/omnihuman-2.0/audio-to-video/v1/01/input-01.png"
    ],
    "audio_url": "https://cdn.vidgo.ai/apis/models/bytedance/omnihuman-2.0/audio-to-video/v1/02/input-02.mp3",
    "prompt": "She speaks warmly to the camera, nodding and gesturing with her right hand. The camera slowly pushes in from a medium shot to a close-up."
  }
}'

Wait for the result

Use the returned task_id to check the same task until it reaches finished or failed. On success, read data.files[].file_url.

Track status

GET https://api.vidgo.ai/api/generate/status/{task_id}

Poll about every 2 seconds while the status is not_started or running. Stop on finished or failed; on success, read data.files[].file_url. For long-running jobs, increase the interval gradually, or add callback_url to the submit request to receive the final result asynchronously.

not_startedrunningfinishedfailed

Submit response

{
  "code": 200,
  "data": {
    "task_id": "8TBISH3H90QQBZ5X",
    "status": "not_started",
    "created_time": "2026-11-05T07:04:47"
  }
}

Finished response

{
  "code": 200,
  "data": {
    "task_id": "8TBISH3H90QQBZ5X",
    "status": "finished",
    "files": [
      {
        "file_url": "https://cdn.vidgo.ai/apis/models/bytedance/omnihuman-2.0/audio-to-video/v2/01/output-01.mp4",
        "file_type": "video"
      }
    ],
    "created_time": "2026-11-05T07:04:47",
    "error_message": null,
    "progress": 100
  }
}

Complete runnable example

Open this after the three-step flow is clear. It combines submission, polling, terminal-state handling, and result extraction in one script.

Complete runnable example

set -euo pipefail
: "${VIDGO_API_KEY:?Set VIDGO_API_KEY in your environment}"

REQUEST_BODY=$(cat <<'JSON'
{
  "model": "bytedance/omnihuman-2.0/audio-to-video",
  "input": {
    "image_urls": [
      "https://cdn.vidgo.ai/apis/models/bytedance/omnihuman-2.0/audio-to-video/v1/01/input-01.png"
    ],
    "audio_url": "https://cdn.vidgo.ai/apis/models/bytedance/omnihuman-2.0/audio-to-video/v1/02/input-02.mp3",
    "prompt": "She speaks warmly to the camera, nodding and gesturing with her right hand. The camera slowly pushes in from a medium shot to a close-up."
  }
}
JSON
)

SUBMIT_RESPONSE=$(curl --silent --show-error --fail-with-body \
  --request POST \
  --url "https://api.vidgo.ai/api/generate/submit" \
  --header "Authorization: Bearer $VIDGO_API_KEY" \
  --header "Content-Type: application/json" \
  --data "$REQUEST_BODY")

TASK_ID=$(printf '%s' "$SUBMIT_RESPONSE" | jq -r '.data.task_id // .task_id // empty')
if [ -z "$TASK_ID" ]; then
  printf 'Submit response did not include task_id:
%s
' "$SUBMIT_RESPONSE" >&2
  exit 1
fi

while true; do
  STATUS_RESPONSE=$(curl --silent --show-error --fail-with-body \
    --url "https://api.vidgo.ai/api/generate/status/$TASK_ID" \
    --header "Authorization: Bearer $VIDGO_API_KEY")
  STATUS=$(printf '%s' "$STATUS_RESPONSE" | jq -r '.data.status // .status // empty')

  case "$STATUS" in
    finished)
      printf '%s' "$STATUS_RESPONSE" | jq -r '(.data.files // .files // [])[]?.file_url'
      break
      ;;
    failed)
      printf '%s' "$STATUS_RESPONSE" | jq -r '.data.error_message // .error_message // "Generation failed"' >&2
      exit 1
      ;;
    not_started|running)
      sleep 2
      ;;
    *)
      printf 'Unexpected task status: %s
' "$STATUS" >&2
      exit 1
      ;;
  esac
done

Request Parameters

model and callback_url belong at the request root; generation parameters belong inside input.

Field	Type	Required	Default	Description
model	string	Yes	—	Fixed to bytedance/omnihuman-2.0/audio-to-video.
callback_url	string (URL)	No	—	A public HTTP(S) URL for receiving task completion notifications.
input	object	Yes	—	Generation parameters: one image in image_urls and one driving audio file in audio_url, plus an optional prompt. Video length follows the audio.
input.image_urls	string[]	Yes	—	Exactly one public HTTP(S) URL of the character or portrait image.
input.audio_url	string	Yes	—	Public HTTP(S) URL of a WAV or MP3 file, up to 15 MB and 4–300 seconds long. The server reads its duration.
input.prompt	string	No	—	Optional guidance for camera movement, actions and scene elements.

Response Fields

Submission returns the task identity immediately. Status responses add progress, output files, or a failure message as the task advances.

Field	Type	Description
code	integer	Application result code. Successful responses use 0 or 200.
message	string	Human-readable response or error message when present.
data.task_id	string	Task identifier used in the status endpoint path.
data.status	string	Current state: not_started, running, finished, or failed.
data.created_time	string	Task creation time in date-time format.
data.progress	number	Completion percentage when the provider reports progress.
data.files[]	array	Generated output files returned after a successful task.
data.files[].file_url	string	Public URL of the generated video.
data.files[].file_type	string	Generated file type, such as video.
data.files[].watermark_url	string \| null	Watermarked output URL when one is available.
data.error_message	string \| null	Failure detail when status is failed.

Task Lifecycle

Treat not_started and running as non-terminal states. finished and failed are terminal alternatives; stop polling when either is returned.

not_started
The task has been accepted and is queued to start.
running
The video is being generated. Continue polling the status endpoint.
finished
The task is complete. Read the video URL from data.files[].file_url.
failed
Generation failed. Read data.error_message for details and stop polling.

Polling and Errors

Authentication401 indicates a missing or invalid Bearer API Key. Check your key in the dashboard before retrying.
Validation400 indicates invalid parameters. Check that image_urls holds one URL and that audio_url is a reachable WAV or MP3 of 4–300 seconds and up to 15 MB.
Polling intervalPoll the status every 2 seconds. For longer tasks, use exponential backoff to gradually increase the interval to 5 seconds.
Terminal statesPoll only while the status is not_started or running. Stop immediately on finished or failed.
Callback optionProvide callback_url to receive an automatic notification when generation finishes, without continuous polling.

Model Specifications

Specification	Value	Description
Model	bytedance/omnihuman-2.0/audio-to-video	Model identifier for asynchronous OmniHuman 2.0 audio-driven video generation.
Input mode	1 image + 1 audio + optional prompt	The audio drives lips, expression and gestures.
Audio	WAV or MP3, 4–300 s, ≤15 MB	Video length follows the audio.
Resolution	720p	Framing follows the input image.
Billing	Hold: audio seconds × 21 credits	Settled against the final charge after completion; the difference is refunded.

View the OmniHuman 2.0 Audio to Video API documentation

OmniHuman 2.0 Audio to Video

OmniHuman 2.0 is ByteDance's audio-driven video model, offered through BytePlus from October 2026. It reads rhythm, intonation and meaning from the audio to drive lip movement, expression and body language for people, animals, cartoons and stylised characters. Video length follows the audio, from 4 seconds up to 5 minutes, which suits presenters, music clips and long narrations.

Why choose the OmniHuman 2.0 Audio to Video API?

Takes of up to 5 minutesDrive a single image with 4–300 seconds of audio; the video runs as long as the track.
Lips, expression and gestures from the audioSpeech and singing set the mouth shapes, facial emotion and body rhythm without manual keyframes.
Prompted camera and actionsUse the optional prompt for camera movement, actions and extra scene elements while the audio handles performance.
Beyond human facesAnimate animals, cartoon and anime characters, 3D figures and multi-person images.
Credits held from the audio lengthAt submission, audio seconds × 21 credits are held; after the task finishes the hold is settled against the final charge and any difference is refunded.

Parameters

Parameter	Requirement	Description
image_urls	Required	Exactly one public HTTP(S) URL of the character or portrait image. Default`—`
audio_url	Required	Public HTTP(S) URL of a WAV or MP3 file, up to 15 MB and 4–300 seconds long. The server reads its duration. Default`—`
prompt	Optional	Optional guidance for camera movement, actions and scene elements. Default`—`

How to Use

Choose the character imageUse one clear image with the face or character visible, hosted at a public HTTP(S) URL.
Prepare the driving audioExport speech or music as WAV or MP3, 4–300 seconds and up to 15 MB.
Add optional directionDescribe camera moves, gestures or added objects in the prompt; leave it empty to let the audio lead.
Submit and collect the videoSend the request to receive a task_id, then poll the status endpoint or use callback_url.

Usage and billing

OmniHuman 2.0 Audio to Video holds credits from the length of the driving audio when you submit, then settles the hold against the final charge for the generated video and refunds any difference.

Usage	Rate	Details
Hold at submission	audio seconds × 21 credits	Reserved from your balance when the task is accepted; for example, 30 seconds of audio holds 630 credits.
Settlement	Final charge	After the task finishes, the hold is settled against the final charge for the generated video and any difference is refunded.

Best Use Cases

Spokesperson and explainer videosTurn a headshot and a recorded script into a presenter clip.
Multilingual presentersPair one image with audio tracks in different languages for localised versions.
Music videos and singing clipsDrive a character with a song for covers, lyric videos and promos.
Course narrationNarrate lessons of up to 5 minutes with a consistent on-screen host.
Talking mascots and petsGive brand mascots, pets or cartoon characters a voice for social posts.

Pro Tips

Use an image where the face is unobstructed and well lit; the mouth area drives the lip sync.
Clean speech without long silences or heavy background noise gives the most readable lip movement.
Use the prompt for what the audio cannot express, such as a slow push-in or the character picking up a cup.
Trim the audio to the moment you need; the hold is calculated from its full length.

Notes

Send exactly one image in image_urls and one WAV or MP3 file in audio_url (4–300 seconds, up to 15 MB).
Video length follows the audio, so the request body carries the image, the audio and the optional prompt only.
Tasks run asynchronously: a successful submission returns a task_id for polling or callback delivery.

Related Models

PixVerse V6 Max Text to VideoPixVerse V6 Max Text to Video turns a prompt into 1–15 second videos at 360p to 1080p, with eight aspect ratios and an optional native audio track for dialogue, effects and music. Describe subjects, actions and sound in one prompt, then pick the length and format for each placement.

PixVerse V6 Max Image to VideoPixVerse V6 Max Image to Video animates one JPEG, PNG or WebP image into a 1–15 second video at 360p to 1080p, with an optional native audio track. The image sets the subject and framing, and a prompt of up to 2,048 characters directs motion, camera and sound.

FLUX 3 Max Text to VideoFLUX 3 Max Text to Video turns a written shot into a 5–20 second 720p or 1080p video with native audio and eight aspect ratios. It follows the camera moves, lighting and sound cues you describe, so one prompt returns picture and audio together.

FLUX 3 Max Image to VideoFLUX 3 Max Image to Video animates a start image into a 5–20 second 720p or 1080p clip with native audio, optionally landing on an end frame you supply. It keeps the composition, subject and palette of your images while adding the motion and sound in the prompt.

FLUX 3 Max Keyframes to VideoFLUX 3 Max Keyframes to Video turns 1–10 images pinned to exact frames into a single 5–20 second 720p or 1080p clip with native audio. Each image appears at the frame_index you set on a 24 fps timeline, so the story beats land where you planned them.

Kling 4.0 Text to VideoKling 4.0 Text to Video turns a prompt or multi-shot script into one 3–30 second 720p or 1080p video with stereo audio and five aspect ratios. It keeps scenes coherent across long takes and shot changes while following your camera, dialogue and sound directions.

Kling 4.0 Image to VideoKling 4.0 Image to Video animates a start image into a 3–30 second 720p or 1080p video with stereo audio, optionally finishing on an end frame you choose. It keeps the subject, framing and look of your image while adding the motion, camera work and sound in your prompt.

Kling 4.0 Reference to VideoKling 4.0 Reference to Video turns reference images, videos and reusable elements into a 3–30 second 720p or 1080p video with stereo audio, cited as @Image1, @Video1 or @Element1. It keeps characters, products and settings consistent with your references inside the new scene you describe.

Kling 4.0 Video EditKling 4.0 Video Edit transforms an existing 3–30 second video from a written instruction, restyling scenes, swapping objects or adding elements at 720p or 1080p. It keeps the original timing, people and camera path that you ask to preserve while applying the change across every frame.

Vidu Q4 Image to VideoVidu Q4 Image to Video turns one start image and an optional motion prompt into a 3–16 second clip at 540p up to 4K. Your image sets the opening frame, so subject, setting and aspect ratio carry through while the prompt directs movement and camera.

Vidu Q4 Reference to VideoVidu Q4 Reference to Video combines a prompt with up to 12 reference images and 3 voice clips into 3–16 second videos at up to 4K. Tag each reference in the prompt to keep characters, products and places recognisable, with optional generated audio at no extra cost.

Gemini Omni 1.1 Flash Text-to-Video preview

Gemini Omni 1.1 Flash Text-to-VideoGemini Omni 1.1 Flash Text-to-Video. Create 4–10 second videos with native audio from text, directing camera, lighting, and sound.

Gemini Omni 1.1 Flash Image-to-Video preview

Gemini Omni 1.1 Flash Image-to-VideoGemini Omni 1.1 Flash Image-to-Video. Animate a start image with native audio and add an optional end frame to guide the closing shot.

Gemini Omni 1.1 Flash Reference-to-Video preview

Gemini Omni 1.1 Flash Reference-to-VideoGemini Omni 1.1 Flash Reference-to-Video. Use 1–7 images to guide characters, products, and style in new video scenes with sound.

Gemini Omni 1.1 Flash Video Edit preview

Gemini Omni 1.1 Flash Video EditGemini Omni 1.1 Flash Video Edit. Edit 3–10 second videos through natural language with up to 5 reference images for visual guidance.

MiniMax H3 Max Turbo Text to Video preview

MiniMax H3 Max Turbo Text to VideoMiniMax H3 Max Turbo Text to Video turns text into coherent 5–15 second videos with synchronized native stereo audio and fast generation. It follows physical motion while rendering subtle changes in light and shadow.

OmniHuman 2.0 Audio to Video API FAQ

What is the OmniHuman 2.0 Audio to Video API?

OmniHuman 2.0 Audio to Video is a ByteDance model for audio-driven video generation from a single image. It turns one character image and 4–300 seconds of speech or music into a 720p video in which lips, expression and gestures follow the audio. The image fixes the character's appearance and framing, while an optional prompt adds camera movement, actions and scene elements. You can call it programmatically or try it from the playground above.

How long can the audio be for OmniHuman 2.0 Audio to Video?

4 to 300 seconds, as a WAV or MP3 file of up to 15 MB. The video length follows the audio, so there is no separate duration setting.

How are credits handled for OmniHuman 2.0 Audio to Video?

When you submit, audio seconds × 21 credits are held from your balance, so 60 seconds of audio holds 1,260 credits. After the task finishes, the hold is settled against the final charge for the generated video and any difference is refunded.

What does the prompt control in OmniHuman 2.0 Audio to Video?

The optional prompt guides camera movement, actions and added scene elements. Lip shapes, emotion and timing come from the audio.

Can OmniHuman 2.0 Audio to Video animate animals or cartoons?

Yes. Besides people it handles animals, cartoon and anime characters, 3D figures and stylised art, as well as images with several people.

What resolution does OmniHuman 2.0 Audio to Video produce?

It outputs 720p video, with framing that follows the input image, so crop the image to the shape you want to publish.

Which audio gives the best OmniHuman 2.0 lip sync?

Clear speech or vocals with little background noise and no long silences. Trim the file to the part you need before uploading.

Can OmniHuman 2.0 Audio to Video sing as well as speak?

Yes. A song or vocal track drives the character's singing, with expression and body movement following the music's rhythm.