AI Talking Avatar v2 API for Expressive Talking Videos

Send one portrait image and one speech track to POST /api/v1/videos/generations with model: ai-talking-avatar-v2. Same request fields and price as ai-talking-avatar, with livelier facial expressions and head movement. The voice is separated from any background music for lip sync, and the original audio is kept in the output. 3 to 60 seconds, 540p or 720p.
Get API Key
Framed portrait on an easel in a softly lit studio beside a microphone

One Photo, a Full Speaking Performance

Livelier expressions and head movement

v2 adds more facial expression and natural head motion than ai-talking-avatar. Use the optional prompt, such as a woman speaking calmly to the camera, to set the tone.
Row of 3D rendered head sculptures turned at different angles

Background music is allowed

The voice is separated from background music before lip sync, and the original audio, music included, stays in the output. Clear speech from one speaker still gives the best result.
Mixing console with two separated channels of light

Output size follows your portrait

The closest of a fixed set of sizes is picked from your portrait's aspect ratio: a 16:9 portrait gives 832×480 at 540p and 1248×736 at 720p; a 9:16 portrait gives 480×832 and 736×1248.
Picture frames of different aspect ratios arranged on a wall
The audio is measured before tokens are deducted; that measurement enforces the 3 to 60 second limit.

AI Talking Avatar v2 Request Fields and Limits

model

Always ai-talking-avatar-v2.

image_url (required)

Portrait of the speaker, up to 10 MB. The face must be clearly visible, unobstructed and roughly front-facing.

audio_url (required)

Speech audio, up to 50 MB, 3 to 60 seconds unless duration is set. Background music is allowed; only one person should be speaking.

resolution

540p (default) or 720p. The 540p tier renders at 480p-class sizes such as 832×480.

duration

Optional integer from 3 to 60. Uses only the first duration seconds of the audio and cannot exceed its length.

prompt

Optional description of the performance. Defaults to person talking.
Same price as ai-talking-avatar, billed per second of output; 720p costs twice as much as 540p.

AI Talking Avatar v2 API Pricing

540p: 0.5 token ($0.01) per second

A 30-second clip costs 15 tokens ($0.30).

720p: 1 token ($0.02) per second

A 30-second clip costs 30 tokens ($0.60).

60-second maximum

The longest output costs 30 tokens ($0.60) at 540p or 60 tokens ($1.20) at 720p.

Billed on measured length

Without duration, the bill uses the measured audio length; with it, exactly duration seconds. Failed jobs are refunded automatically.

Integrate the AI Talking Avatar v2 API

1

1. Create an API key

Generate a key at /keys and send it as Authorization: Bearer <key> from your server.
2

2. Prepare the inputs

Pick a front-facing portrait with the aspect ratio you want the video to have, and a speech track from one speaker.
3

3. Submit the job

POST {"model": "ai-talking-avatar-v2", "image_url": "https://…", "audio_url": "https://…", "resolution": "720p"} to /api/v1/videos/generations.
4

4. Collect the video

Store img_uuid, poll GET /api/v1/jobs/detail or wait for your callback, and download the result. 720p takes roughly three times as long as 540p.
Create an API Key
Talking clips from a single photo and a voice line.

What Developers Build with AI Talking Avatar v2

Spokesperson and explainer videos

Turn one headshot into announcements and product explainers driven by recordings or text-to-speech.

Narrated content with music beds

Use voiceovers that already have background music; the voice is separated for lip sync and the full mix is kept.

Course and training narration

Give lesson narration a presenter face without filming each lesson.

Multilingual versions

Pair translated audio with the same portrait to produce localized talking clips.
Inputs, differences from v1, length and pricing.

AI Talking Avatar v2 API FAQ

What is the AI Talking Avatar v2 API endpoint?

POST https://api.goenhance.ai/api/v1/videos/generations with model: ai-talking-avatar-v2.

How is it different from ai-talking-avatar?

The request fields and price are the same. v2 gives livelier facial expressions and head movement and separates the voice from background music before lip sync.

What size is the output?

It follows your portrait's aspect ratio. A 16:9 portrait gives 832×480 at 540p and 1248×736 at 720p.

Can the audio contain music?

Yes. The voice is separated out for lip sync and the original audio, music included, stays in the video. Only one person should be speaking.

How long can the output be?

3 to 60 seconds. With duration set, the audio itself may be longer than 60 seconds.

How is it priced?

0.5 token ($0.01) per output second at 540p and 1 token ($0.02) per second at 720p.

Start Building with the AI Talking Avatar v2 API