ElevenLabs Launches Dubbing v2

ElevenLabs has launched Dubbing v2, an innovative AI dubbing model that preserves the original speaker's emotion and performance as is. This model goes beyond simple text translation, conveying the speaker's tone, pace, delivery, and emotional intent across 90+ languages.

Existing AI dubbing systems relied on text-based translation, often missing the speaker's subtle nuances. Dubbing v2, on the other hand, directly conditions on the original performance, transferring the speaker's inflection and energy into other languages to produce much more natural results.

Key features include the following:

• Sync-aware translation: Translated sentences are automatically synchronized to the start, end, and pace of the original video, delivering professional dubbing quality.
• Workflow automation: It automates the costly, specialized dubbing process that previously required translators, voice actors, and engineers, increasing accessibility for creators and businesses.

This service offers tailored solutions depending on the target audience. Creators can easily localize YouTube content, while marketers can scale global campaigns with a consistent emotional tone. Additionally, studios and broadcasters can build professional localization workflows through ElevenProductions.

It is currently available through ElevenCreative and ElevenProductions, with API support planned for the future.