API Reference

Talking Image ​

POST https://api.newportai.com/api/async/talking_image

Talking Image API could generate a talking video from audio and an image, including non-human facial images such as pets or animated characters.The talking image is capable of mirroring the facial expressions and vocal tone of a real person，while supporting multiple languages.

  • Use Talking Image API to get taskId with POST method.
  • Use Polling API to get Talking Image result with POST method.

Request Headers ​

  • Content-TypestringRequired
    Enum
    application/json
  • AuthorizationstringRequired
    Bearer authentication format, for example: Bearer

Request Body ​

  • audioUrlstringRequired
    Original audio url. The audio format supports mp3, wav and m4a. we recommend using m4a.
  • photoUrlstringRequired
    Image url. Maximum resolution: 1280 * 1280.

Response ​

  • codeinteger
    0 success, or error code.
  • messagestring
    "success", or error msg.
  • dataobject
    taskIdstring
    Use the taskId to request the Polling API to retrieve the generated outputs.

Example ​

Request:

zsh
curl -X POST 'https://api.newportai.com/api/async/talking_image' \
-H "Authorization: Bearer {{key}}" \
-H 'Content-Type: application/json' \
-d '{
  "audioUrl": "https://dreamface-resource-new.s3.amazonaws.com/youtube/music/youtube-2t6hJEQ3Ubs.m4a",
  "photoUrl": "https://dreamface-resource-new.s3.amazonaws.com/demo.jpg"
}'

Response:

json
{
  "code": 0,
  "message": "success", 
  "data": {
    "taskId": "b08337dc08d7428daa64b3d5e61b8350"
  }
}

Next Step: ​

Use Polling API to get the result.

Demo ​