Elevenlabs announced the language expansion of its latest artificial intelligence (AI) Text-to-Spicha (TTS) model last week. With this expansion, the AI model now supports 41 new languages, the total calculation takes to 70 supported languages. The New York City -based AI startup stated that with this expansion, the model is now accessible to 90 percent of the global population. In particular, the company released the eleven V3 (alpha) model on 8 June, and pighed it as the “most expressive TTS model”.
Eleven V3 now supports 70 languages
One in Post On X (East was known as Twitter), the official handle of XIlabs announced that her latest AI model, XI V3, now supports additional 41 languages. With this update, the model can originally produce audio from the text script in a total of 70 languages. Some of the newly added languages include Arabic, Assamese, Bengali, Bulgarian, Catalan, Gujarati, Latvian, Malay, Malayalam, Marathi, Nepali, Swahili, Tamil and Telugu.
The company advised that people wishing to generate lessons in any new language should record an instant voice clone (IVC) while choosing the language. In addition, XIbabs are also adding voice library voice to new languages in the coming weeks.
Eleven V3 is the successor of multilingual V2 and V2.5 TTS models. Latest AI Model Supports Inline audio tags such as whispering, excited, ah, and more. Adding audio tags enables the model to add expressive emotional nuances to the audio generation, non-verbal signals, and dramatic distribution.
It also supports multi-speaker interaction with obstruction, natural pacing and overlapping dialogues. Additionally, the company says that the model handles elements such as better stress, rhythm and relevant awareness. Eleven V3 is available through the company’s website and mobile app. It is currently not available as application programming interface (API).
In April, Elevenlabs introduced a new enterprise-centric agentic feature Dub agent transfer. Part of the company’s connivance AI, it gives two AI agents to interact and share the conversation. This feature creates a system where an AI agent can hand over the conversation with the other, more special agent with conversation data.