Generate AI Avatar Videos

Quality of the lip-sync: depending on whether you choose to “train a personalized avatar” (in the Add Avatar and Train Lip-sync section), the system will automatically apply either the universal lip-sync model or the dedicated trained lip-sync model for your avatar. If you skip “train a personalized avatar,” the lip-sync results for your avatar will be less realistic compared to using a dedicated lip-sync model trained with your avatar’s video.

Get the id of the avatar as "_id" returned by /api/v1/anchor/character_list.

Why is video generation sometimes slow? Please pay special attention to the parameter isSkipRs, which corresponds to Smart Motion Adjustment in the web UI. Enabling this option enhances the naturalness of the AI avatar’s movements. However, it can significantly slow down the generation process.

Request

Authorization

Provide your bearer token in the

Authorization

header when making requests to protected resources.

Example:

Authorization: Bearer ********************

Body Params application/json

title

string

optional

The title of video

anchor_id

string

required

The "_id" returned from the api/v1/anchor/character_list. When you call character_list, remember to set user_video_twin_id and type.

anchor_type

integer

required

Type of the avatar. 0 is system provided. 1 is user uploaded (custom)

back_id

string

optional

ID of the system default background image. Do not set this value if you do not want to replace background.

custom_back_id

string

optional

ID of the custom background image. Do not set this value if you do not want to replace background

color

string

optional

RGB format string of the color of the custom background image. e.g. rgb(133,137,132)

audioSrc

string

required

The url of the audio. Get the audio file url from send_tts method. Alternatively, Iif you provide your own audio. You can record the audio by yourself. You can obtain the audio from other TTS vendor such as ElevenLabs.

resolution

integer

required

Resolution of the video

web_bg_width

integer

required

Width of the video. Simply set this to 0 if you do not need background matting (e.g. Neither back_id nor custom_back_id is not provide)

web_bg_height

integer

required

Height of the video. Simply set this to 0 if you do not need background matting (e.g. Neither back_id nor custom_back_id is not provide).

web_people_width

integer

required

The placement of the avatar within the final image. Width of the avatar. Simply set this to 0 if you do not need background matting (e.g. Neither back_id nor custom_back_id is not provide)

web_people_height

integer

required

The placement of the avatar within the final image. Height of the avatar. Simply set this to 0 if you do not need background matting (e.g. Neither back_id nor custom_back_id is not provide)

web_people_x

integer

required

The placement of the avatar in the final video. The x value of the upper left corner of the avatar base，relative to the upper left corner of the video. Simply set this to 0 if you do not need background matting (e.g. Neither back_id nor custom_back_id is not provide)

web_people_y

integer

required

The placement of the avatar in the final video. The y value of the upper left corner of the avatar base，relative to the upper left corner of the video.

isSkipRs

boolean

optional

Skip intelligent action adjustment (default true). If you turn this on, your video generation will be significantly slowed down.

anchor_background_img

string

optional

The url of the original background image corresponding to the avatar, used to matting.
No need to set anchor_background_color if this value is given.

anchor_background_color

string

optional

RGB format string of the original background color corresponding to the avatar, used to matting. e.g. rgb(0,255,0)
No need to set anchor_background_img if this value is given.

isAllowReverse

string

optional

Allow Reverse. True is allow.

isToPublicPool

boolean

optional

(Only useful for dedicated line users)
Whether to send your video generation request to the public computing power pool for inference, true means yes.
As a dedicated line user, there is no monthly video generation limit set by us. Your capacity is determined by how many videos your dedicated lines can process within a month. This endpoint allows you to send video generation requests to the public computing pool when your dedicated lines are occupied. For example, if you have two dedicated lines but three video generation requests are created simultaneously, the third request would typically wait until one of the lines becomes available. However, if you choose to utilize the public pool, all three requests can be processed concurrently, ensuring faster results and increased efficiency.

The requests sent to the public computing pool will consume your coins as regular user's requests. The purpose of sending your requests to the public computing pool is to accelerate the process of your video generation by using more computing resources. But this will increase your cost (you need to prepare enough coins).

isCaptionEnabled

boolean

optional

Whether to enable captions, true means enabled.

msg

string

optional

This field contains the audio text content, must provide a valid value when generating captionAlign

captionAlign

object

optional

This is the font style, which is only effective when isCaptionEnabled equals true.

language

string

optional

Caption Language.

Example:

en-US

PrimaryColour

string

optional

RGBA format string of the font primary colour.

Example:

rgba(146, 84, 222, 1)

FontName

string

optional

Font name,
Currently supported fonts are Arial, AaYuanWeiTuSi, AaFengKuangYuanShiRen, Aozora Mincho, SJbangkaijianti, 337-SSHuaGuangTi, YEFONTZhaoPaiTi, ziticqnananti, ZXFMingKeBen, TangXianBinSong, Luoguocheng-maobixiaoxiankaijia, Kingnamype Yuanmo SC, JMH, huangkaihuaLawyerfont.

Example:

AaFengKuangYuanShiRen

Fontsize

integer

optional

Font Size.

Example:

BorderStyle

integer

optional

Border Style.

Example:

subtitle_position

string

optional

Subtitle Position,It represents the percentage distance of the font from the bottom of the video, displayed in center alignment.

Example:

0.3

OutlineColour

string

optional

RGBA format string of the font outline colour.

Example:

rgba(0, 0, 0, 1)

BackColour

string

optional

RGBA format string of the font back colour.

Example:

rgba(146, 84, 222, 1)

Example

{
  "title": "My first AI video",
  "anchor_id": "660556b3cd492168141926d3",
  "anchor_type": 1,
  "back_id": "62a6d3e56c209514b155b12b",
  "audioSrc": "https://speech_cdn.ai-anchor.com/dubbing_stable/pyK1dvqcDKpKJo1IcEPm_9714_fr.mp3",
  "resolution": 1080,
  "web_bg_width": 853,
  "web_bg_height": 480,
  "web_people_width": 270,
  "web_people_height": 480,
  "web_people_x": 292,
  "web_people_y": 0,
  "isSkipRs": true,
  "isCaptionEnabled": true,
  "captionAlign": {
    "language": "en-US",
    "PrimaryColour": "rgba(247, 89, 171, 1)",
    "OutlineColour": "rgba(0, 0, 0, 1)",
    "BorderStyle": 4,
    "BackColour": "rgba(146, 84, 222, 1)",
    "FontName": "AaFengKuangYuanShiRen",
    "Fontsize": 50,
    "subtitle_position": 0.3
  }
}

Request samples

Shell

JavaScript

Java

Swift

PHP

Python

HTTP

Objective-C

Ruby

OCaml

Dart

Responses

🟢200generate

application/json

Body

0 is OK

code

integer

required

data

object

required

title

string

required

msg

string

required

algo_result

string

required

result

string

required

retry

integer

required

process

integer

required

isDel

boolean

required

isSkipRs

boolean

required

status

string

required

error

string

required

anchor_type

string

required

resolution

string

required

voice_type

string

required

audioSrc

string

required

custom_voice

string

required

priority

integer

required

restart_date

null

required

audioLen

integer

required

web_people_width

integer

required

web_people_height

integer

required

web_people_x

integer

required

web_people_y

integer

required

web_bg_width

integer

required

web_bg_height

integer

required

color

string

required

result_cover

string

deprecated

wl_model

string

required

anchor_background_img

string

required

anchor_background_color

string

required

_id

string

required

The id of the video. e.g. 652d00b8a71c91a83bba6195

user_id

string

required

anchor_id

string

required

web_dist_bg_width

integer

required

web_dist_bg_height

integer

required

web_dist_bg_x

integer

required

web_dist_bg_y

integer

required

replace_bg_id

string

required

version

integer

required

createdAt

string

required

updatedAt

string

required

__v

integer

required

Example

{
  "code": 0,
  "data": {
    "title": "202404071749",
    "msg": "",
    "algo_result": "",
    "result": "",
    "retry": 0,
    "process": 0,
    "isDel": false,
    "isSkipRs": false,
    "status": "start",
    "error": "",
    "anchor_type": "1",
    "resolution": "1080",
    "voice_type": "0",
    "audioSrc": "https://dh24as48lv9ce.cloudfront.net/ai/speech/tts/azure_service/azure_synth.-8161556380964742732.wav",
    "custom_voice": "",
    "priority": 4,
    "restart_date": null,
    "audioLen": 5,
    "web_people_width": 608,
    "web_people_height": 1080,
    "web_people_x": 657,
    "web_people_y": 0,
    "web_bg_width": 1920,
    "web_bg_height": 1080,
    "color": "",
    "result_cover": "",
    "wl_model": "WLGPNormal",
    "anchor_background_img": "",
    "anchor_background_color": "",
    "_id": "6613b4c216004c1df435d6ff",
    "user_id": "637498294abbb165572d5dbe",
    "anchor_id": "660556b3cd492168141926d3",
    "web_dist_bg_width": -2,
    "web_dist_bg_height": -2,
    "web_dist_bg_x": 0,
    "web_dist_bg_y": 0,
    "replace_bg_id": "62a6d3e56c209514b155b12b",
    "version": 1,
    "createdAt": "2024-04-08T09:11:30.329Z",
    "updatedAt": "2024-04-08T09:11:30.329Z",
    "__v": 0
  }
}