HeyGen Avatar V (Direct)
HeyGen Avatar V rendered through HeyGen's own API with Wireflow's account key. It is the only node that can use avatars trained into Wireflow's HeyGen account. Drive it with an audio track; output is always 16:9 1080p.
HeyGen Avatar V (Direct)
Badge: New
Node type: video:heygen_avatar_v
Category: video
Description
HeyGen Avatar V rendered through HeyGen's own API with Wireflow's account key. It is the only node that can use avatars trained into Wireflow's HeyGen account. Drive it with an audio track; output is always 16:9 1080p.
Pricing
- Cost: 10 credits per second of audio
- Minimum: 30 credits (anything under 3 seconds bills as 3)
- Maximum: 60 seconds of audio per render
Avatar V is billed per second because HeyGen bills us per second (~$0.0667/s), and the render comes out exactly as long as the audio you feed it. A 12-second track costs 120 credits; a 45-second one costs 450.
Audio longer than 60 seconds is refused before you are charged rather than truncated. Split the script into shorter takes and join the clips downstream.
The duration is measured from the audio file itself, on the server, before any credits move. If the file's duration can't be read (an unusual container, or a host that won't serve the file), the run is refused and you are not charged. Use wav or mp3 from an upstream node.
The credit figure shown on the canvas is an estimate: it uses the length of the connected audio when the editor has probed it, and a 10-second example otherwise. The real charge always uses the measured length.
Canvas ports
These appear as port handles on the left side of the node.
| ID | Label | Details |
|---|---|---|
audio_url |
Audio | AUDIO (required) · The speech track that drives the avatar. Avatar V is audio-driven, so there is no text/voice path on this node. |
Sidebar config
These render as form fields in the right-side config panel when the node is selected.
| ID | Label | Details |
|---|---|---|
avatar_id |
Avatar | TEXT (required) · An avatar id from Wireflow's HeyGen account. Pick one from the Avatar list. Pasting a GROUP id is rejected before you are charged. |
motion_prompt |
Motion Prompt | TEXT · multiline · Optional direction for gesture/motion, e.g. "speaks calmly, hands mostly still". |
remove_background |
Remove Background | BOOLEAN |
Outputs
| ID | Label | Type |
|---|---|---|
video |
Generated Video | VIDEO |
What makes this node different
The other HeyGen nodes (HeyGen Avatar (Digital Twin), HeyGen Avatar4) run
through FAL, and FAL calls HeyGen on FAL's account. That means they can
only see HeyGen's public stock catalog, so a trained/custom avatar is
unreachable there, and those nodes refuse a custom id before charging you.
This node calls api.heygen.com directly with Wireflow's own account key, so
the avatars trained into Wireflow's HeyGen account are the ones you pick
from. The Avatar list in the node settings is that account's directory.
Who can use which avatars
Trained avatars are real people's likenesses, so for now they are admin-only. Everyone else gets HeyGen's stock catalog, which is what the Avatar list shows. The restriction is enforced server-side before any credits are deducted, so pasting a trained avatar's id into the field does not work around it.
Fixed settings
- Aspect ratio is 16:9 and is not configurable. Avatar V letterboxes anything else (a 9:16 request comes back as a pillarboxed 16:9 frame), so the node pins 16:9 rather than shipping a control that silently degrades the output. Crop to vertical downstream if you need 9:16.
- Resolution is 1080p.
- Output format is mp4.
How to wire it
Avatar V is audio-driven, so there is no script/voice field. Connect an
audio source to the Audio port (an ElevenLabs TTS node, an audio input node,
or anything that outputs an audio URL). The avatar's mouth follows that track.
Audio uploads are capped at 32MB by HeyGen, and at 60 seconds by this node; longer scripts should be split into separate takes.
Avatar ids vs group ids
HeyGen has two kinds of 32-character id and they look identical:
- an avatar (look) id, which is what this node needs
- an avatar group id, a container holding several looks
Picking from the Avatar list always gives you a look id. If a group id ends up in the field, the node refuses the run before any credits are deducted and tells you which look ids that group contains.
Auto-generated from the Wireflow node registry.