Lip Sync for Video Projects
Lip sync is a new workflow stage for video projects. After the AI voice-over, it re-animates the speakers' lips to match the translated…
AI voice-over now uses ElevenLabs v4, the newest and most natural ElevenLabs voice model, by default. Every new project in every workspace and region uses v4, whether the file arrives through the UI, API, templates, Storyline (.story) export, or an integration. The model is set to Auto, and Auto now means v4, so nothing needs to be switched on. The price stays the same at 10 Smartwords per word.
v4 keeps the speaker's voice consistent across lines and regenerated lines, reads numbers and brand names such as "S&P 500" or "0.01%" in the language of the sentence, and speaks with a natural accent in the target language even when the voice was recorded in another language. It supports 90+ languages.
Existing projects keep their current model, so nothing a customer has already approved changes. To get v4 for existing content, create a new project, or select ElevenLabs v4 in the editor under Audio → TTS model. The file is then voiced again and charged as a new dub.
To learn more in our Pronunciation Control for AI Voiceover help center article.