Text to talking avatar

Turn any avatar into a talking video

Pick a face you made — or upload a photo — write up to 400 characters, and choose a voice. We handle the speech and the animation.

12 credits per second 5 s minimum 30 s maximum Final cost is based on the actual audio length.

1. Choose a face

Private by default

New accounts get 30 free credits. You can also upload a photo after signing in.

Example faces

Examples are for preview only. Generate your own on the home page after signing in.

2. Write the script

0 / 400

≈ 5 s Short sentences sound better.

3. Pick a voice

8 voices

Estimated ≈ 5 s · 60 credits

final cost is based on the actual audio length

Typical render: 1–3 minutes. You can leave this page while it finishes.

Made with AvatarSay

A face, a script, a voice

The scripts below are the actual prompts used. No voice cloning, no editing trickery.

Face: a friendly fox astronaut with a tiny helmet

“Hey there! I'm Captain Fox. Today we're flying to the moon — and yes, I packed snacks.”

Voice: Friendly Person

Face: a cheerful barista with freckles and a messy bun

“Welcome back to the channel! Grab a coffee, because this one is a long story.”

Voice: Lively Girl

Face: a smiling man in a white shirt, soft studio light

“Hi, I'm Daniel. I help small teams ship better products. Let's talk.”

Voice: Casual Guy

How it works

Three steps, no studio

We synthesize the speech first, measure the real audio, then animate the face. That is why the final cost can differ slightly from the estimate.

Pick the face

Use an avatar you generated or upload a clear, front-facing photo. One person, eyes visible, mouth unobstructed.

Write and choose

Keep it conversational. Pick one of eight voices and preview it before you commit.

We render it

The job runs in the background. Watch it on this page or close the tab and find the result in your Dashboard.

Good to know

A few honest limits

Up to 30 seconds. Longer scripts are not supported in this version.
One person, front-facing. The model works best with a clear, unobstructed face.
No real public figures. Do not animate celebrities, politicians or anyone else without consent.
AI-generated. Disclose it where the platform or your audience expects it.

Talking avatar FAQ

Before you hit render

Still stuck? support@avatarsay.com reaches a person, usually within one business day.

How long does a talking avatar take?

Most videos finish in one to three minutes. The page polls every four seconds, and the job keeps running if you close the tab. You can come back to it from your Dashboard.

What does a video cost?

Talking video is 12 credits per second with a 5-second minimum. A 5-second clip costs 12 × 5 credits. The final charge is based on the actual audio length, not the estimate shown before you submit.

Can I use my own photo?

Yes. Upload a PNG, JPG or WebP up to 8 MB and tick the consent box confirming it is you, or that you have the person’s permission. Uploads are private and are never used to train models. Read the Acceptable Use Policy before uploading.

What makes a good source image?

One person, facing the camera, with the whole face visible and an unobstructed mouth. Even light and a plain background help. Group photos, heavy shadows, sunglasses and extreme side profiles usually produce a worse result.

How long can the script be?

Up to 400 characters, which is roughly 30 seconds of speech. Very short scripts are padded to the 5-second minimum.

What happens if the render fails?

We refund the credits automatically and record the refund in your credit history. The page will tell you the exact amount. If a refund does not appear, email support@avatarsay.com with the video link or time.