Upload one portrait and the speech you want. The AI talking photo maker turns it into a video of that face saying every word.
You get an MP4 as long as your audio. Most short clips take 3–5 minutes.
One of our own tests: a single portrait and one line of speech, turned into a talking video. Turn the sound on.

The line he says: “Hi there. A minute ago this clip was silent. Now every word I say matches my lips.”
Standard tier · 6.5 s of speech · 56 credits · about 3.5 minutes
We made the same man say the same line twice: once with the AI talking photo, once by animating the photo and then lip syncing it.
Photo plus audio in, a speaking video out. One step, and the head and face move with the speech.
56 credits
First an 8-second silent clip from the photo, then the mouth redrawn to the line. Two steps, and you pick the motion in the first one.
30 + 64 = 94 credits
For a single speaking portrait, the AI talking photo is cheaper and simpler. Animate first when you need a specific movement, a longer shot or a scene around the face.
An AI talking photo is a video made from one still image and an audio track: the face in the photo opens its mouth, blinks and moves its head in time with the speech. You give it a portrait and a voice recording, and it renders every frame of that person talking.
It gives a face to a voice without a camera. A product explainer, a course intro, a family photo or an AI character can speak in minutes instead of a filming day.
The model reads the audio and works out the mouth shape for each sound, plus the small moves that come with speech: a nod on a stressed word, raised brows on a question, blinks between phrases. It then draws each frame from your photo, keeping the face, hair and clothes the same while the expression changes.
The person's identity, hairstyle, clothing, background and the framing of your photo.
Lip movement matched to the words, natural blinks, small head and shoulder motion, and your audio as the soundtrack.
The whole AI talking photo runs in your browser. No camera, no editing software.
A JPG or PNG of one face looking at the camera, at least 300 pixels on each side.
An MP3, WAV, M4A or AAC recording. Up to 15 seconds on Standard and HD, 30 seconds on Cinematic.
Choose the quality, tick the consent box and press Animate. Most short clips are ready in three to five minutes.
Every tier uses the same photo and audio. They differ in sharpness, how long the speech can be and how lifelike the motion is.
8 credits per second
Social posts, drafts and quick tests. Our example above is Standard.
16 credits per second
Presenters, ads and anything shown full screen on a desktop.
27 credits per second
Longer speeches, expressive delivery and the most lifelike head and body motion.
The AI talking photo can only animate what it can see. These five things make the biggest difference.
Front-facing or slightly turned. A side profile shows only half a mouth.
The model needs both corners of the lips to shape each word.
A neutral or soft smile gives the model room to open the mouth naturally.
A wide grin or open mouth in the photo fights every word.
Even light on the face, no heavy shadows, and at least 300 pixels on each side.
Detail around the eyes and lips is what makes the motion believable.
Crop group photos down to the one face you want to speak.
With several faces, the wrong one may start talking.
No hands, microphones, sunglasses or hair across the mouth.
Covered features are redrawn badly or not at all.
The video is exactly as long as your audio, and you pay for those seconds. A little preparation saves credits.
Cut quiet seconds at the start and end. They cost the same as speech.
Record in a quiet room, close to the microphone. Background music or a second voice can make the lips follow the wrong sound.
Short breaks between sentences give the face time to blink and reset, which looks more human.
The mouth follows the sounds, not the words, so any spoken language or accent is fine.
Three ways to get a person on screen saying your words. Pick by what you start with.
Price on Imgveo
40 credits for 5 s on Standard
Price on Imgveo
40 credits for a 5 s clip
Price on Imgveo
From 30 credits for an 8 s clip with sound
Prices are read from our live price list.
Anywhere a face would say it better than text on a slide.
Give a product page, a course or an onboarding flow a friendly face that speaks.
Use the same portrait for every video so viewers recognise the presenter.
Let a generated character introduce itself, tell a story or answer a fan.
Generate the character facing the camera with a closed mouth, like our example.
Bring an old portrait to life with a message recorded by a relative.
Scan the photo sharply and crop it to one face before you upload.
Record the same message in several languages and use one portrait for all of them.
Keep each recording a similar length so the videos feel consistent.
Honest limits, so you know what to expect before you spend credits.
The face, head and shoulders move. Walking, gestures and camera moves need a video model instead.
If half the mouth is hidden, the motion looks wrong. Use a front-facing photo.
Standard and HD take up to 15 seconds, Cinematic up to 30. Split longer scripts into several clips.
You need the right to use the face and the voice. We block content that breaks our rules.
What the AI talking photo accepts, what it returns and what it costs.
Photo: JPG or PNG up to 10 MB, at least 300 px, no wider or taller than 2.5 to 1. Audio: MP3, WAV, M4A or AAC up to 10 MB.
An MP4 as long as your audio, with your speech as the soundtrack, at 720p on Standard and 1080p on HD and Cinematic.
8, 16 or 27 credits per second of audio, rounded to whole seconds. Paid plans only. Failed runs are refunded.
| Audio length | Standard | HD | Cinematic |
|---|---|---|---|
| 5 s | 40 | 80 | 135 |
| 10 s | 80 | 160 | 270 |
| 15 s | 120 | 240 | 405 |
| 30 s | — | — | 810 |
A talking photo can make anyone appear to say anything, so we ask for permission first. Before every run you confirm that you have the right to use the face and the voice, and every photo is screened.
Go deeper: guides from our blog
Open Imgveo Studio to run this and every other AI video and image model from a single workspace, with one credit balance.
Give it a voice, then take it further.
Already have a video? Sync its mouth to new speech instead.
Lip sync a videoAnimate a photo with camera moves and a scene around it.
Animate a photoTake your talking photo to 2K or 4K for big screens.
Upscale a videoGenerate new scenes with dialogue and sound built in.
Try Veo 3.1Upload a portrait and a voice track, and the AI talking photo maker does the rest in a few minutes.