ElevenLabs Dubbing (v1 and v2)
Dub audio or video into English, Spanish, German, French, Japanese and more with ElevenLabs Dubbing. It keeps each speaker's voice and timing, preserves background audio, and returns subtitles.
## Overview
ElevenLabs Dubbing translates spoken audio in a recording into another language. It keeps each speaker's tone, pacing, and timing. It also keeps the original background audio, so mixes often need less rework. The output helps you ship localized content without re-recording voice talent.
## What you can build
- Localized YouTube videos, course lessons, and product demos
- Multi-speaker podcast translation with preserved speaker identity
- International versions of interviews and documentaries
- Fast language variants for ads and social clips
- Subtitles in the target language timed to the original speech
## Inputs
- An audio or video file to dub.
- A target language: English, German, French, Turkish, Chinese, Korean, Japanese, Hindi, Italian, or Spanish.
- An optional source language. Auto detects it when you don't know it.
- An optional cloning strength from 0 to 10. Higher values sound closer to the original speaker.
- An optional list of key terms to keep as they are, such as names and brands.
## Outputs
- An audio input returns the dubbed audio as MP3.
- A video input returns the same video with the dubbed voice.
- Subtitles in the target language as a WebVTT file.
## Recommended settings
- Use eleven_dubbing_v2 for more natural, expressive dubs, and eleven_dubbing_v1 for lower cost.
- Raise cloning strength when voice match matters more than natural delivery.
- Lower cloning strength for language pairs with very different phonetics.
- Add key terms for product names, people, and places. Keep each term short.
## Limitations
- Multi-channel audio isn't preserved.
- Noisy audio, heavy music over speech, and low-bitrate recordings can reduce accuracy.
- Unclear speaker separation and overlapping dialogue can cause speaker mix-ups.
## Safety & compliance
- Only dub content you have rights to use and translate.
- Don't use dubbing to mislead people about what someone said.
- Get consent before cloning a real person's voice into another language.
API quick start
Run dubbing with a single API call.
{
"model": "eleven_dubbing_v1",
"targetLanguage": "en",
"inputAudio": "https://your-cdn.com/input.mp3",
"inputVideo": "https://your-cdn.com/input.mp4"
}curl -X POST "https://api.wiro.ai/v1/Run/elevenlabs/dubbing" \
-H "Content-Type: application/json" \
-H "x-api-key: YOUR_WIRO_API_KEY" \
--data-binary @- <<'JSON'
{
"model": "eleven_dubbing_v1",
"targetLanguage": "en",
"inputAudio": "https://your-cdn.com/input.mp3",
"inputVideo": "https://your-cdn.com/input.mp4"
}
JSONOther models you might like
Explore similar audio & speech models.



