Generate a digital human video

Generate lip-synced digital human videos by uploading driver videos and audio files.

Step 1: Submit Generation Task

Endpoint: POST https://api.navtalk.ai/generate

Request Example:

curl -X POST "https://api.navtalk.ai/generate" \
  -H "Content-Type: application/json" \
  -d '{
    "License: YOUR_LICENSE_KEY",
    "video_url": "https://your-domain.com/driver_video.mp4",
    "audio_url": "https://your-domain.com/target_audio.mp3"
  }'

📌 Parameter Description:

Parameter name	Type	Explanation
video_url	string	Please provide a video link, preferably one that clearly shows the face and mouth movements
audio_url	string	Input audio link in MP3/WAV format

✅ Successful Response:

{
  "status": "started",
  "task_id": "14cb760f-05ac-4fd3-a82c-e841f2f005d0"
}

🔹Step 2: Query Task Status and Results

Use the returned task_id to check the processing results:

curl -X GET "https://api.navtalk.ai/query_status?license=YOUR_LICENSE&task_id=14cb760f-05ac-4fd3-a82c-e841f2f005d0"

✅ Successful Response:

{
    "code": 200,
    "message": "SUCCESS",
    "data": {
        "resultUrl": "https://easyaistorageaccount.blob.core.windows.net/easyai/uploadFiles/2025/07/03/2749bdb2-2220-4315-acd1-1783a45eaac6.mp4",
        "state": "done"
    }
}

📌 Status Description:

Status Value	meaning
started	The task has been created and is currently being processed
processing	In video compositing
done	Successfully completed; the results can be downloaded
failed	The synthesis has failed. You can retry or check the error message.

Notes and Recommendations

• The interface response time is usually between 5 to 20 seconds, depending on the file size and server load.

• It is recommended to use the task_id polling method at intervals of 2 to 3 seconds to check the status.

📘 For more parameters and format descriptions, please refer to the Digital Human API documentation.