Avatar Video
Make any face talk
One still image and a line of text. It does not care whether the face is a person, a character, an animal, or a painting.
no subscription
- Over 120 voices
- Or clone your own
- No watermark
- Commercial use
Avatar Video
If it has a face, it can talk
Most talking-head tools want a person, a studio recording and a subscription. This takes one still image and a line of text, and it does not care whether the face is real. A finished minute costs about $3, and a month you make nothing costs nothing.
And a voice to go with it
No studio
No camera, no recording, no shoot day
The subscription tools want a person who can sit still under a light, or a stock presenter who is also presenting for everyone else. This wants a picture. The face can be one that exists, one you drew, or one that was painted three hundred years ago.
Nothing is trained and nothing is waited on: the same still can say something different an hour later, in another voice, in another language.
- 01Send one still imageA headshot, a character render, a photo of your dog, a painting. One picture with a face in it is the whole input.
- 02Give it something to sayType a line and pick from over 120 studio voices, clone your own from about thirty seconds of audio, or upload a recording you already have.
- 03Get the videoA talking MP4 up to 90 seconds, at 720p or 1080p. Download it and drop it anywhere.
The alternative
A year of talking-head video
Billed in the months you shipped nothing, the months you were busy, and the month you meant to cancel.
Charged only where you made something. A quiet month costs nothing, and there is nothing to cancel when it is over.
Per finished minute the cheapest subscription tiers are competitive, and if you publish every week one of them may well suit you better. This is for everyone whose year does not look like that: the months are the argument, not the rate.
Try it
Five dollars to see it work
Nothing recurring, nothing to cancel. $0.75 per started 15 seconds at 720p and $1.13 at 1080p, which is about $3 for a finished minute.
- 1080p, up to 90 seconds
- 120 voices, or clone your own
- No watermark, ever
- Full commercial rights
- Bring your own audio instead
- Balance never expires
- Unused balance is refundable
- Failed renders are refunded
Once you have a balance
Make them in your browser
Sign in, do it on screen, download the file. Nothing to install and nothing to write.

Or without opening it
Two ways for people who would rather not
A weekly update is a script and a face, which is what code is good at, so the same video is one line of it, or one sentence to an assistant like Claude.
Ask an assistant like Claude
youHey Claude, have my avatar read this week’s update
- list_voicespicks a voice from the catalog
- upload_filesends the face
- generate_avatar_videosays the line, in sync
A written example. The tool names are the real ones and the file is a real render, but this is not a recording of a session.
Or call it yourself
curl -X POST "https://api.neurall.io/v3/generations" \
-H "Authorization: nl-YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"feature": "video/avatar",
"args": {
"image": "f_9c21e0b4",
"voiceId": "aria",
"text": "Whatever you have a picture of, I can make it talk.",
"resolution": "720p"
},
"variants": {
"resolution": "720p"
}
}'One POST starts the generation and a GET tells you when it has landed. Same account, same balance, same price as the website.
More than this
Neurall makes a lot more than this
It is all in the same place, on one balance, and everything you make is yours to sell.
Questions, answered
What do I actually upload?
One still image with a face in it, and either a line of text or an audio file. That is the entire input: no studio recording, no green screen, no camera, and nothing to train.
Does the face have to be a real person?
No, and that is the part that makes this different. A drawn character, a 3D render, an animal, a mascot, a painting that has never moved: if there is a face in the picture, it can speak. Most talking-head tools are built around human presenters and quietly assume one.
How much is it, exactly?
$0.75 per started 15 seconds at 720p, and $1.13 at 1080p. So a finished minute is about $3, or $4.50 at 1080p. It is billed per started block, so a 20 second video costs two blocks. Nothing recurring.
How long can one video be?
Up to 90 seconds per video. For anything longer, make it in parts and cut them together, which is also how you would want to shoot it.
Where does the voice come from?
Over 120 studio voices are included, in the language you write in. You can also clone your own voice from about thirty seconds of audio, and then it reads anything you write.
Can I use audio I already have?
Yes. If you have a recording, upload it instead of a script and the face is animated to that, which is the route to use when the performance matters more than the convenience.
Can I use it commercially?
Yes. Every video you generate is yours, with no watermark, no attribution and full commercial rights. Consent still applies: the Terms of Use require that you have a person’s permission before you generate their likeness.
Is there a free trial?
No. Every clip on this page is a real render made through the product, so you can see and hear exactly what you would be buying before spending anything. You add balance and pay per video, and whatever you do not spend is refundable for 60 days.
Give it a face and a line.
Add funds, generate, own the result. No subscription, no watermark, you pay per generation.
















