Lipsync

Pricing

Choose the model with the model field. Defaults to sync-lipsync-2.

sync-lipsync-2       credits = max(50,  ceil(seconds) x 10)
sync-lipsync-2-pro   credits = max(100, ceil(seconds) x 20)
sync-lipsync-3       credits = max(150, ceil(seconds) x 30)

Duration is measured from your video automatically before processing, so you do not need to send it.

Maximum video duration: 30 seconds.

When to use sync-lipsync-2-pro

sync-lipsync-2 rebuilds the mouth region at a fixed internal resolution and composites it back into your frame, re-encoding as it goes. On tightly framed or vertical video the result can look soft next to the surrounding footage.

sync-lipsync-2-pro runs a super-resolution pass over that region and preserves the bitrate of your source. Measured on a real 480x854 clip: the source arrived at 6.5 Mbps, sync-lipsync-2 returned 3.1 Mbps, sync-lipsync-2-pro returned 6.7 Mbps. Use it when the face is small in frame, when you can see pixelation around the mouth, or when the output will be judged on detail.

For talking-head footage where the face fills the frame, sync-lipsync-2 is usually indistinguishable and half the price.

When to use sync-lipsync-3

The newest model. It handles obstructions in front of the mouth, close-ups, and extreme head angles without special configuration — cases where the other two struggle.

No model can add resolution your source never had. A 480x854 input stays 480x854.

Credits are deducted upfront based on the measured duration. Refunded automatically if lipsync fails.

Multi-person video

active_speaker (boolean, default false) turns on active-speaker detection. When more than one face is visible, it works out who is actually talking and drives only that mouth. Leave it off for single-subject video.

----------|---------|-------------------|---------------------| | 3s | 50 | $0.30 | $0.40 | | 5s | 50 | $0.30 | $0.40 | | 10s | 100 | $0.60 | $0.80 | | 15s | 150 | $0.90 | $1.20 | | 15.04s | 160 | $0.96 | $1.28 | | 20s | 200 | $1.20 | $1.60 | | 30s | 300 | $1.80 | $2.40 |

Maximum video duration: 30 seconds.

Credits are deducted upfront based on the measured duration. Refunded automatically if lipsync fails.

Endpoint

POST https://vicsee.com/api/v1/tools/lipsync

See Authentication for API key setup.

Parameters

Sync Modes

Example Request

curl -X POST https://vicsee.com/api/v1/tools/lipsync \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "video_url": "https://example.com/scene.mp4",
    "audio_url": "https://example.com/voiceover.mp3"
  }'

Response

Lipsync returns 202 Accepted immediately with a task id. It does not wait for the job.

{
  "success": true,
  "data": {
    "id": "task_abc123",
    "model": "sync-lipsync-2",
    "status": "pending",
    "pollUrl": "/api/v1/tasks/task_abc123",
    "duration": 5.0,
    "creditsUsed": 50,
    "creditsRemaining": 450,
    "createdAt": "2026-03-01T12:00:00.000Z"
  }
}

duration and creditsUsed are returned up front so you can reconcile the charge without waiting. The video is not in this response.

Collecting the result

Poll GET /api/v1/tasks/{id} until status is completed or failed. This is the same task endpoint every other VicSee generation uses.

{
  "success": true,
  "data": {
    "id": "task_abc123",
    "model": "sync-lipsync-2",
    "status": "completed",
    "result": {
      "url": "https://cdn.vicsee.com/results/dew/user-id/abc123.mp4",
      "type": "video",
      "duration": 5.0
    }
  }
}

Jobs typically take 30 to 250 seconds depending on length and model. Poll every few seconds. Result URLs are hosted on the VicSee CDN and available for 7 days.

Pipeline Example

Generate a voiceover with ElevenLabs, then lip-sync it to an AI-generated video:

# Step 1: Generate voiceover
AUDIO=$(curl -s -X POST https://vicsee.com/api/v1/generate \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "elevenlabs-v3-dialogue",
    "prompt": "Welcome to the future of AI video creation."
  }' | jq -r '.data.output.url')

# Step 2: Lip-sync the voiceover to your video
curl -X POST https://vicsee.com/api/v1/tools/lipsync \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d "{
    \"video_url\": \"https://example.com/ai-generated-scene.mp4\",
    \"audio_url\": \"$AUDIO\"
  }"

Limits

Errors

Related

• ElevenLabs Audio — Generate voiceover audio to lip-sync
• Merge Audio + Video — Merge audio and video without lipsync
• Merge Videos — Combine multiple videos into one
• Tasks API — Check task history
• Credits API — Check your credit balance