Last week, ElevenLabs revealed the expansion of its artificial intelligence (AI) text-to-speech (TTS) model to include an additional 41 languages. This update brings the total to 70 supported languages, which the company claims allows the AI model to reach 90 percent of the global population. The New York City-based startup launched the Eleven V3 (alpha) model on June 8, calling it their “most expressive TTS model” to date.
Eleven V3 Now Supports 70 Languages
The announcement was made via a post on X (formerly Twitter), where ElevenLabs confirmed the addition of 41 new languages to their AI model, Eleven V3. This update allows the model to generate audio from text in a total of 70 languages, including Arabic, Assamese, Bengali, Bulgarian, Catalan, Gujarati, Latvian, Malay, Malayalam, Marathi, Nepali, Swahili, Tamil, and Telugu.
For those interested in generating content in any of the new languages, the company has recommended recording an Instant Voice Clone (IVC) while choosing the desired language. In the upcoming weeks, ElevenLabs will also introduce Voice Library voices for the newly supported languages.
Eleven V3 builds on the features of its predecessors, the multilingual V2 and V2.5 TTS models. The latest model offers inline audio tags, enabling users to incorporate elements such as whispers, excitement, sighs, and other expressive nuances. These audio tags enhance the model’s ability to convey emotional depth, non-verbal communication cues, and dynamic delivery in generated audio.
Additionally, Eleven V3 facilitates multi-speaker interactions, allowing for natural pacing, interruptions, and overlapping dialogues. The company emphasizes that the model has improved capabilities in managing stress, cadence, and contextual awareness. It can be accessed via the company’s website and mobile applications, although it is currently unavailable as an application programming interface (API).
Earlier in April, ElevenLabs unveiled a new feature designed for enterprise use, known as Agent Transfer. This addition to the company’s Conversational AI suite enables two AI agents to engage in dialogue and share conversation details, allowing for a seamless transition of a conversation from one specialized agent to another.
