If the job is replacing or localizing a spoken track, start with ElevenLabs Multilingual V2.
That answer changes once the dub is no longer one speaker carrying one script across languages, but a real multi-speaker scene.
This Page Solves A Narrow Audio-Localization Problem
This guide follows Rivya's live spoken-audio catalog as it stood on April 21, 2026.
- public paths cross-checked:
/audio,/ai-models, and current live voice-model pages - related product guides reviewed: Audio Workflows in Rivya, Video Workflows in Rivya, and References and Uploads in Rivya
- this page is only about spoken-track replacement and localization inside Rivya's audio area, not full lip-synced video localization
The useful split here is not "what language?"
It is "what kind of spoken replacement is this?"
The Three Dubbing Paths
| Dubbing job | Best first path | Why it fits |
|---|---|---|
| one speaker, same script across languages | ElevenLabs Multilingual V2 | the best path when language transfer is the main job |
| one speaker, one straightforward replacement | ElevenLabs Turbo 2.5 | faster when the job is simpler than "dubbing" makes it sound |
| several speakers in one dubbed scene | ElevenLabs Dialogue V3 | better when turn-taking and role structure matter |
Those paths are adjacent, but they are not interchangeable.
Choose By Replacement Shape
Use ElevenLabs Multilingual V2 when the heart of the workflow is carrying one spoken delivery into another language while keeping it coherent.
Use ElevenLabs Turbo 2.5 when the task is only a simple spoken replacement: one speaker, one language, one clean re-read.
Use ElevenLabs Dialogue V3 when the output is not one replaced line, but a whole spoken scene with several voices.
That is the fastest way to keep dubbing from drifting into the wrong branch.
What This Page Does Not Promise
In Rivya today, "dubbing" is most accurately the spoken-audio replacement and localization layer.
This page is not promising:
- full lip-synced video translation
- timeline-level video editing
- automatic mouth-match correction on uploaded footage
If the real requirement is a finished video where motion and audio need to land together, the better paired pages are AI Voiceover for Videos, AI Video Generator With Audio, and Video Workflows in Rivya.
A Faster Dubbing Decision Order
If you want the shortest reliable order, use this:
- decide whether the output is one speaker across languages, one simple replacement, or a multi-speaker dubbed scene
- if language transfer is the hard part, start with ElevenLabs Multilingual V2
- if the job is only one clean replacement, start with ElevenLabs Turbo 2.5
- if the script behaves like a scene, switch to ElevenLabs Dialogue V3
Where To Go Next
- If the real task is one-speaker narration, read AI Narration Generator.
- If the real task is video-specific voice-over, read AI Voiceover for Videos.
- If the real task is a generated clip with native audio, read AI Video Generator With Audio.
- If you need the broader workflow guides, read Audio Workflows in Rivya, Video Workflows in Rivya, and References and Uploads in Rivya.
Prepare A Dubbing Brief In Rivya
Before the first run, make the replacement job explicit:
- Source: decide whether you are replacing an existing spoken track or creating a localized read from text.
- Speaker shape: mark whether this is one speaker, one simple replacement, or a multi-speaker scene.
- Language notes: include target language, names, product terms, pronunciation, and pacing constraints.
- Output use: decide whether the draft is for review, publishable audio, video handoff, or a localization batch.
The first useful run should prove that the dubbing shape is correct before you expand into more languages or more voices.
Review The Dub As A Replacement
Listen for whether the target-language audio still carries the same message, role, pronunciation, and timing.
If the language transfer is wrong, revise the language and pronunciation notes. If the speaker structure is wrong, switch branches before adding more style words. If the draft is close, save it in History before making language variants.
