DomoAIUpload the portrait
Use a clear, front-facing photo or character render. One reference image can be attached per generation, and the face should be unobstructed for the cleanest result.
Workflow: talking avatar
DomoAI Talking Avatar: Make a Portrait Speak
Give a portrait a voice. Pair a photo, illustration, or character render with a script or an audio track and DomoAI matches mouth shapes and head movement to the audio, so a still image delivers a line instead of sitting frozen.
Audio to mouth shapesScript or voice trackHead movement included720p and 1080p output
Three Steps from Input to Finished Clip
Every DomoAI workflow starts from a prompt and an optional reference file. Nothing has to be rebuilt by hand.
Add the audio
Attach a voice track up to 15 seconds per generation, or write the script and let the workflow handle the delivery. Lip sync auto-match keeps audio and mouth shapes aligned.
Generate the take
Render the clip and check the sync against the audio, then download at 720p or 1080p and upscale to 4K when the take is final.
How This Workflow Compares
DomoAI workflows differ by the material you already have. Pick the one that matches it.
| Workflow | What you provide | Good for |
|---|---|---|
| Talking avatar | A portrait plus audio or a script | Making a still portrait deliver a line with matched mouth shapes |
| Character to video | A character image plus a motion clip | Making a character perform a physical action rather than speak |
| Image to video | A photo, illustration, or product render | Adding ambient motion and camera moves to a still |
| Video to video | An existing clip plus a style reference | Restyling footage you already shot |
Questions About This Workflow
Answers about inputs, limits, resolution, and how credits are used.
What is a DomoAI talking avatar?+
It is a short clip built from one still portrait and an audio track. DomoAI animates the face so the mouth shapes and head movement follow the audio you supplied.
Can I use a script instead of a recording?+
Yes. You can attach a voice track or write the script and let the workflow handle the delivery, then check the sync before exporting.
How long can a talking avatar clip be?+
One audio track of up to 15 seconds can be attached per generation, and generated clips run from 5 to 60 seconds depending on the workflow.
What kind of portrait works best?+
A clear, front-facing image with an unobstructed face. Illustrations and character renders work as well as photos, as long as the mouth area is visible.
Start With the Material You Already Have
Every new account starts with 100 welcome credits, valid for 30 days, so you can test this workflow before choosing a plan.