Each clip is one video plus one new voice track. Turn the sound on to check the mouth, or hit Try this example to load the same video and audio.
AI lip sync takes a video of someone talking and a different audio track, and redraws the mouth so it matches the new words. The lips, jaw and teeth follow every sound of the new audio, frame by frame, so the person looks like they really said it.
Everything else in your lip sync video stays as it was: the face, the head moves, the background and the camera. Only the mouth area changes, and the old sound is replaced with your new track. That makes it the quickest way to fix a line, dub a clip into another language or give a character a new voice without filming again.
Add an MP4 or MOV of 2 to 10 seconds with one face clearly turned to the camera. Longer clips are cut to the first 10 seconds.
Upload an MP3, WAV or M4A of the words you want said: your own recording, a dub in another language, or speech made from text with the AI Voice Generator.
Press Create. In a few minutes the mouth moves to the new audio, and the new audio is the soundtrack. Watch it in the browser and download the MP4.
Upload a recording, a voice note or a text-to-speech file. The mouth follows the sound, and your audio is used as is.
Lip sync follows the sound, not the text, so a dub in Spanish, Hindi or Japanese lines up just like English.
A 5-second lip sync costs 15 credits, the same as the free credits every new account gets. No card needed for the first one.
Same face, same moves, same framing. The video comes out in 720p with the shape and length of the clip you uploaded.
A front or slightly turned face gives the cleanest result. Side profiles and faces looking away are hard to sync.
The AI syncs one face. With several people, crop the clip to the one you want to talk first.
Keep hands, mics and hair away from the lips. Clear, well lit footage keeps the teeth and lips sharp.
Short clips of one shot work best. Clips with cuts, fast camera moves or a face that leaves the frame can glitch.
A clear voice with little background noise gives the sharpest mouth shapes. Loud music under the voice can blur them.
The whole track drives one mouth. If two people talk, the face will say both parts.
Aim for audio about as long as the video. Extra audio is cut off at the end of the video, and a shorter track leaves a silent ending.
Write the line in the AI Voice Generator, pick a voice or clone your own, download the MP3 and upload it here.
The result always has the same length as your video. Here is what happens to the audio.
| Your files | What you get |
|---|---|
| Audio and video about the same length | The mouth follows every word from start to end. This gives the best result. |
| Audio longer than the video | The video plays once and the audio is cut where the video ends. Trim the audio first so no words are lost. |
| Audio shorter than the video | The person speaks the audio, then the rest of the video plays with no sound and a closed mouth. |
| Video longer than 10 seconds | Only the first 10 seconds are used. Cut your clip to the part you want before uploading. |
Three ways to make a face say your audio on ImagesArt. Pick by what you start with.
Translate the script, record or generate the new voice, and the speaker's mouth matches the new language.
Swap a wrong price, a name or a whole sentence in a clip you already shot, without filming again.
Make a cartoon, a game character or a friend (who said yes) say a funny line for TikTok, Reels or Shorts.
Test several scripts on one creator clip and keep the version that sells best.
With several people in frame, one face is changed. Crop to the speaker first.
Head moves, hands and background stay as filmed. To make new motion, use Image to Video.
You bring the new audio. There is no script box here; make speech from text with the AI Voice Generator first.
Upload your own footage, or footage of someone who said yes. Do not make anyone appear to say things they never said.
AI lip sync costs 3 credits per second of video, counted in 5-second blocks: 15 credits for a clip up to 5 seconds, 30 credits for 6 to 10 seconds. The audio is free. New accounts get 15 free credits, enough for one 5-second lip sync. Failed runs are not charged.
Cancel anytime · Keep access until the end of your billing period
Upload a video and a new voice track, and get a lip-synced clip in a few minutes.