heygen/avatar-iv

Create realistic talking avatar videos from text with HeyGen's Avatar IV engine

Capabilities

1:1 4:3 3:4 16:9 9:16 2:3 3:2

Community model (estimated from hardware time)

Name	Type	Description	Default	Constraints
`avatar_id` *	string	Unique identifier of the avatar. Get available avatar IDs from the HeyGen List All Avatars API.	`—`	—
`input_text` *	string	Text that the avatar will speak. Must be less than 5000 characters.	`—`	—
`voice_id` *	string	Unique identifier of the voice. Get available voice IDs from the HeyGen List All Voices API.	`—`	—
`avatar_style`	string	Visual style of the avatar.	`"normal"`	normal closeUp circle
`caption`	boolean	Whether to enable captions in the video.	`false`	—
`height`	integer	Height of the output video in pixels.	`1080`	min: 100, max: 2160
`title`	string	Title of the video.	`""`	—
`voice_emotion`	string	Adds emotion to voice, if supported by the selected voice. Set to 'none' for no emotion.	`"none"`	none Excited Friendly Serious Soothing Broadcaster
`voice_speed`	number	Voice speed. Value ranges from 0.5 to 1.5.	`1`	min: 0.5, max: 1.5
`width`	integer	Width of the output video in pixels.	`1920`	min: 100, max: 3840

avatar_id required string

Unique identifier of the avatar. Get available avatar IDs from the HeyGen List All Avatars API.

input_text required string

Text that the avatar will speak. Must be less than 5000 characters.

voice_id required string

Unique identifier of the voice. Get available voice IDs from the HeyGen List All Voices API.

avatar_style string

Visual style of the avatar.

Default: "normal"

normal closeUp circle

caption boolean

Whether to enable captions in the video.

Default: false

height integer

Height of the output video in pixels.

Default: 1080 min: 100, max: 2160

title string

Title of the video.

Default: ""

voice_emotion string

Adds emotion to voice, if supported by the selected voice. Set to 'none' for no emotion.

Default: "none"

none Excited Friendly Serious Soothing Broadcaster

voice_speed number

Voice speed. Value ranges from 0.5 to 1.5.

Default: 1 min: 0.5, max: 1.5

width integer

Width of the output video in pixels.

Default: 1920 min: 100, max: 3840

Version: 47140d7dd37b Updated: 6/26/2026 508 runs