AI talking video maker that makes any character talk.
Upload a picture of a character, type what they say and pick a voice, or upload your own recording. Clippie animates them saying it, with the voice in time with their lips, in up to 30 seconds.
- Clip length4 to 30 s
- Shapes5
- QualityUp to 1080p
Talking clips made from one picture and one line each, with the voice lined up with the lips. Tap a clip to hear it.
How to make a picture talk. Four steps.
- 1
Add your character
Upload a picture or pick one from your library. A face looking toward the camera works best: a mascot, a 3D character, an animal, a drawing.
- 2
Give them a line
Type what they say and pick a voice, adding tags like [whispers] or [laughs] for acting. Or upload a recording of one person speaking, up to 29 seconds.
- 3
Pick a shape and quality
9:16, 16:9, 1:1, 3:4 or 4:3, at 480p, 720p or 1080p. The price shows before you generate.
- 4
Generate
It takes two to four minutes. The clip lands in your library, ready to download or drop into a video.
A talking clip from one picture. No filming, no rigging.
Lips in sync with the voice
The character is animated saying the line, and Clippie lines the voice you picked up with the lips, so the character still sounds like themselves.
Type it or upload it
Generate the voice from a line, or upload your own recording and the character speaks in your voice.
Voice acting tags
Write [whispers], [laughs], [sighs] or [nervous] in a line and the voice performs it. The Voice acting button adds tags for you without changing a word.
Thousands of voices
125 hand-picked voices plus a search of thousands more in the voice library.
Five shapes
9:16 for TikTok, Reels and Shorts, 16:9 for YouTube, plus 1:1, 3:4 and 4:3 for posts and ads.
Pay for the real length
A typed line is priced on a long estimate and the difference comes back once the voice is made. Failed clips are refunded.
Frequently asked questions.
What is an AI talking video generator?
A tool that turns a still picture into a video of that character speaking. In Clippie you upload a picture, type a line and pick a voice (or upload a recording), and Clippie animates the character saying it, with the voice lined up with the lips. Clips run 4 to 30 seconds.
Do the lips really match the voice?
Yes. The character is animated saying the line, then Clippie transcribes both the animation and your voice line and stretches the voice onto the mouth movements, word by word. You hear the voice you picked, in time with the lips. If a clip can’t be lined up, Clippie tries again, and a clip that fails is refunded.
Can I use my own voice?
Yes. Choose “Upload audio” and add a recording of one person speaking with no music, up to 29 seconds (mp3, wav, m4a, aac or ogg). The clip runs as long as the recording and your voice is lined up with the character’s lips.
What pictures work best?
One character with their face toward the camera, well lit, not too small in the frame. 3D characters, cartoon mascots, animals and illustrations all work. To make a new character first, draw one in AI Studio’s image models or in AI Story’s character dialog.
How do I make two characters talk to each other?
Use Characters talk in AI Story. Each scene is one character saying one line in their own voice, and “Make them talk” runs every scene through the same talking model, so a whole conversation comes out as one video with captions.
Can I use a photo of a real person?
Only a photo you have the right to use, such as one of yourself, and you confirm that when you upload it. In AI Story a photo can be used as it is or redrawn in the story’s look. Pictures and lines that break the content rules are refused, and you are not charged.
How much does it cost?
Talking clips use AI credits from your Clippie plan or a credit top-up, priced by the clip’s length and quality (480p or 720p, and 1080p on the Talking page). You see the price of every clip before it starts. A typed line is charged on a generous estimate and the difference comes back once the voice is made, and a clip that fails is refunded. Exporting the finished video uses your plan’s export minutes.
