HeyGen Avatar V API

Render talking-avatar videos from a script or audio track with a HeyGen photo avatar or digital twin

Modalities
Image, Text, VideoVideo
Price
$0.12
Resolution
Up to 1080p
Released
Nov 3, 2026

HeyGen Avatar V overview

HeyGen Avatar V by HeyGen

HeyGen Avatar V is HeyGen's most advanced avatar model and its highest-fidelity engine. HeyGen designed it to preserve a person's identity, motion, gestures, expressions and mannerisms across generated videos.

HeyGen Avatar V specs

Model ID
heygen/avatar-v
Developer
HeyGen
Input
Image, Text, Video
Output
Video
Endpoint
POST https://api.aimlapi.com/v2/video/generations

Built around a reference video

HeyGen describes Avatar V as a Diffusion Transformer with flow matching. It conditions directly on the full token sequence of the reference video rather than on bottleneck embeddings. Sparse Reference Attention keeps the computational cost almost linear with the length of that reference.

HeyGen says a short recording of about 15 seconds can serve as the reference for studio-quality videos with lifelike motion, multi-angle stability and long-form performance. Treat that length as the starting point for the source recording.

Set the frame and the resolution

Aspect ratio accepts 16:9, 9:16, 1:1, 4:5, 5:4 or auto. Auto keeps the aspect ratio of the source and falls back to 16:9 when the source cannot be read. Start with auto, then set an explicit ratio only when the output must differ from the source.

Resolution accepts 720p or 1080p. Start with 1080p, the higher of the two values.

Avatar V or Avatar IV

HeyGen recommends Avatar V primarily for real-human avatars and digital twins. For virtual, 2D, 3D and non-human characters it recommends Avatar IV, which adds control over gestures and expressiveness.

HeyGen says Avatar V supports digital-twin, studio-avatar and photo-avatar looks when the individual look lists Avatar V among its supported engines. Pick a look that carries that listing.

Use cases

Render a digital twin of a founder delivering a product update from a written script.

Turn a recorded audio track into a talking-avatar video of a real-human avatar.

Produce vertical 9:16 clips of the same avatar for short-form social feeds.

Create 16:9 training videos at 1080p with the same on-screen presenter in every lesson.

HeyGen Avatar V API pricing

HeyGen Avatar V pricing in US dollars
UsagePriceUnit
Video generation$0.12

Prices from the AI/ML API price list, checked on Nov 12, 2026.

How to use the HeyGen Avatar V API

View docs

Build with your agent

Give your agent the model details, code and documentation.

1Get your API key

Create API key
export AIMLAPI_KEY="8f2b4c7d9a1e6b3f…"

2Make your first request

curl 'https://api.aimlapi.com/v2/video/generations' \
  -H "Authorization: Bearer $AIMLAPI_KEY" \
  -H 'Content-Type: application/json' \
  -d '{
  "model": "heygen/avatar-v",
  "avatar_id": null
}'

The request returns a generation id. Poll GET https://api.aimlapi.com/v2/video/generations?generation_id=<id> until the status is completed.

HeyGen Avatar V FAQ

How much does HeyGen Avatar V cost?

HeyGen Avatar V costs $0.12 / sec on AI/ML API.

What inputs does HeyGen Avatar V take?

It takes image, text and video inputs and returns a video. The speech comes from a script or an audio track.

What does the auto aspect ratio do?

Auto keeps the aspect ratio of the source. If the source ratio cannot be read, the output is 16:9.

Which avatar looks work with Avatar V?

HeyGen says digital-twin, studio-avatar and photo-avatar looks work when the individual look lists Avatar V among its supported engines.

When should I use Avatar IV instead of Avatar V?

HeyGen recommends Avatar IV for virtual, 2D, 3D and non-human characters. It recommends Avatar V primarily for real-human avatars and digital twins.

What is the HeyGen Avatar V model ID?

Use heygen/avatar-v as the model parameter. Requests go to https://api.aimlapi.com/v2/video/generations with your AI/ML API key.

Start building with HeyGen Avatar V

HeyGen Avatar V API

heygen/avatar-v

Build with your agent

Give your agent the model details, code and documentation.

1Get your API key

Create API key
export AIMLAPI_KEY="8f2b4c7d9a1e6b3f…"

2Make your first request

curl 'https://api.aimlapi.com/v2/video/generations' \
  -H "Authorization: Bearer $AIMLAPI_KEY" \
  -H 'Content-Type: application/json' \
  -d '{
  "model": "heygen/avatar-v",
  "avatar_id": null
}'