Log inUpgrade

Dubbing · 6 min read

How to dub a video into another language with lip sync

Open Lip Sync Studio, upload your video (MP4 or MOV) and paste the translated line in the Script tab. Pick one of 21 ElevenLabs voices, or upload a recording of the line in the Voice tab, then tap Generate: you get the same clip with the mouth re-synced to the new words. Kling LipSync takes clips of 2–10 seconds at 720p or 1080p, and Sync Lipsync 3 takes up to 60 seconds. The studio doesn’t translate, the voices aren’t the person’s own, and you sign in to generate.

Your video
MP4 or MOV under 100 MB
Length
2–10 seconds on Kling LipSync; up to 60 seconds on Sync Lipsync 3 and the HeyGen models
New words
A script up to 400 characters, read by one of 21 ElevenLabs voices, or your own recording
What changes
The mouth, to match the new words; the face, head movement and background stay
Sign-in
Required to generate

The workflow

  1. Upload the video you want to dub

    Open Lip Sync Studio, tap Upload a photo or video and choose an MP4 or MOV under 100 MB. One person facing the camera, mouth in view, works best. The studio picks the model from your clip: Kling LipSync for 2–10 seconds with each side 720–1920 pixels, otherwise Sync Lipsync 3, up to 60 seconds. Both sit under Video to video in the Model menu, with HeyGen Lipsync and HeyGen Lipsync Precision. Use video of yourself or of someone who agreed.

  2. Paste the line in the new language

    In the Script tab, paste the translated line under “What should they say?”, up to 400 characters, in any language. The studio doesn’t translate, so translate it first and have a native speaker check it. Keep it about as long to say as the original. Kling LipSync keeps your whole clip: a shorter line leaves the original mouth moving with no sound at the end, and a longer one is cut where the clip ends.

  3. Pick a voice, or use a real one

    Below the line, the Voice menu holds 21 ElevenLabs voices; Hear it plays a short English sample of the one picked. The voice is made from your line when you generate. These are stock voices, so the dub won’t sound like the person in the video. For a real voice, switch to the Voice tab beside Script and tap Upload audio or Record: the translated line, said by you or by someone who agreed. It takes MP3, WAV, M4A or the sound of a video, 2–60 seconds.

  4. Generate, then check it with sound

    Tap Generate; the button shows the price. If you aren’t signed in, you sign in now and your video, words and settings wait for you. The studio uploads the clip, makes the voice first and checks the exact price, then the video appears under Generations. Watch it with sound: every word, the lips on p, b and m, the teeth up close, and the last seconds of the clip.

Recorded Made demonstration

An English podcast intro, dubbed into Spanish

This is a real run from November 12, 2026. Theo is one of Lip Sync Studio’s preset faces, made by Made, not a real person, and his 8-second English clip was made by Made too. The Spanish line was typed in the Script tab and read by Brian, one of the ElevenLabs voices, and Kling LipSync, the model the studio picks for this clip, re-synced his mouth to it. The requests are the studio’s own, sent through Made’s API from a script rather than tapped through the browser. A speech-to-text check heard the Spanish line word for word, and the lips close on its p and m sounds. Up close, the teeth are a smooth white strip, and the mouth stays half open in the short pause between the two sentences. The Spanish ends about a second and a half before the clip does, and for that stretch you see his original mouth, a small movement and a closing smile, with no sound. The new voice replaces the clip’s sound, and no native speaker has judged the accent. It’s one run, not a measure of how often a dub holds.

Original clip, in English
Dubbed into Spanish

Kling LipSync, with an ElevenLabs Turbo v2.5 voice · Generated · View the example record and price basis (JSON)

Prompts you can inspect and adapt

Recorded generation prompt

What the voice read for this run

Bienvenidos de nuevo al programa. Hoy vamos a convertir una sola foto en la introducción de un episodio completo.

This is the line as typed under “What should they say?”: the studio sends it to the voice word for word. Kling LipSync takes no prompt of its own, only your clip and that voice.

Prompt recipe to adapt

Ask for a translation that fits the clip

Translate this line into natural spoken Spanish for a lip-synced dub: "[paste your line]". Keep the meaning and the tone, and make it about as long to say as the original, around [8] seconds. Use everyday words, no stage directions and no emoji. Reply with the line only.

Use it with a translator or an assistant you trust, then paste the line under “What should they say?”. Fill in the language and your clip’s length, and have a native speaker check the line before you post.

How a dub is priced

By the model and the seconds it bills, plus the voice when you use the Script tab. The total appears on Generate before you run it.

Kling LipSync bills your clip’s length in started 5-second blocks. Sync Lipsync 3 bills the voice’s length, up to your clip’s, and the HeyGen models bill the voice’s length. In the Script tab, the voice is priced by the line’s characters and added to the total.

Price source: Kling LipSync on the Made price sheet · Checked

The studio checks the exact price after upload; if it came out higher than the button showed, it shows the new price and waits for another tap. If the provider it was priced on fails, a backup provider can take the run at its own price, which can be higher. Failed videos aren’t charged, and neither is a voice that doesn’t finish.

What to review before you use the result

What you seeWhat to try
The mouth keeps moving after the voice endsKling LipSync keeps your whole clip, so when the new line is shorter, the rest shows the original mouth moving with no sound. Write a line that fills the clip, or trim the clip to the line before you upload it.
The end of the line is cut offKling LipSync and Sync Lipsync 3 stop where your clip ends, and the studio warns you when the voice runs longer. Shorten the line or use a longer clip.
The accent sounds wrongThe voices are stock ElevenLabs voices, and in another language they can keep an English accent. Try another voice, or record a native speaker saying the line and upload it in the Voice tab.
Your clip is refusedUse an MP4 or MOV under 100 MB. Kling LipSync takes 2–10 seconds with each side 720–1920 pixels; Sync Lipsync 3 takes up to 60 seconds with each side 256–3840 pixels. Dub a longer video in parts.
HeyGen Lipsync says the lengths don’t matchThe HeyGen models need the clip about as long as the voice, and the studio shows the range that works. Trim the clip, or use Kling LipSync or Sync Lipsync 3.
The music or room sound is goneWith Kling LipSync, the finished clip carries only the new voice. Add music back in your editor, or mix it into the track you upload in the Voice tab.

Questions about this workflow

Does Lip Sync Studio translate the video for me?

No. It doesn’t transcribe or translate. Paste the line already in the new language, or upload a recording of it, and the studio re-syncs the mouth to those words.

Will the dub sound like the person in the video?

Not with a script. The studio reads it in one of 21 stock ElevenLabs voices and doesn’t copy anyone’s voice. To keep a real voice, record the translated line yourself, or have someone who agreed record it, and upload it in the Voice tab.

Which languages work?

The script box takes any language, and we checked two. A speech-to-text check heard our Spanish line word for word and our Hindi one nearly so, but some Hindi sounds kept an English accent. Try a short line first in any other language, and for a native accent, upload a native speaker’s recording.

How long a video can I dub?

Kling LipSync takes clips of 2 to 10 seconds; Sync Lipsync 3, HeyGen Lipsync and HeyGen Lipsync Precision take up to 60. A voice track can run 2 to 60 seconds. Dub a longer video in parts.

Do I need to sign in?

Yes, to generate. You can upload the clip, write the line and pick the voice first; when you tap Generate and sign in, the studio keeps them for you.

Work with Claude and Made

Through the Made connector, Claude can do the same two steps: make the voice from your translated line with ElevenLabs Turbo v2.5, then lip-sync your uploaded clip to it with Kling LipSync, quoting each price first. It doesn’t run Lip Sync Studio’s checks, so tell it the clip’s length and keep the line about as long.

Conversation starter

Ask your assistant

Use Made to dub my attached 8-second video into Spanish. First make the voice with ElevenLabs Turbo v2.5 and the voice Brian, reading exactly: "Bienvenidos de nuevo al programa. Hoy vamos a convertir una sola foto en la introducción de un episodio completo." Then lip-sync my video to that voice with Kling LipSync. Show me both prices before you run anything.

If your client can’t upload a video, use Lip Sync Studio with the same clip and line: it picks the model and checks the lengths for you.

Keep creating