ElevenLabs has launched its new text-to-speech model, Eleven V3, now supporting 70 languages, including several Indian languages. This model can generate emotionally expressive, natural-sounding voices.
ElevenLabs, a leading company in artificial intelligence-based voice technology, has made a significant leap with its new text-to-speech (TTS) model, Eleven V3. The company announced that its new version now supports 41 new languages, bringing the total to 70. This update allows the model to connect with approximately 90 percent of the world's population through voice technology.
Significant Support for Indian Languages
Among the 41 newly added languages by ElevenLabs are several Indian languages, which is significant news for Indian users. These include Hindi, Assamese, Bengali, Gujarati, Malayalam, Marathi, Nepali, Tamil, and Telugu. This is expected to greatly expand the reach of this technology in a linguistically diverse country like India.
Information Shared on Social Media
ElevenLabs announced via a post on its official X (formerly Twitter) account that Eleven V3 can now convert text to speech in a total of 70 languages. This means users can now type text in their preferred or native language and hear it spoken in a natural and emotionally expressive voice.
Recommendation for Instant Voice Clone (IVC)
The company also advises users who want to generate content in a new language to use the Instant Voice Clone (IVC) feature for that language. This allows users to provide a sample of their own voice or another voice to get output in a similar style.
Furthermore, the company stated that in the coming weeks, it will add Voice Library Voices for these new languages, giving users the option of pre-recorded voices.
Features of the New Technology
The Eleven V3 model is an advanced version of its previous multilingual models, V2 and V2.5. This new model includes several special features:
- Emotional Audio Tags: Sounds like whispers, sighs, enthusiasm, and disappointment can now be added to AI voices.
- Multi-Speaker Support: This model better represents real-life conversations with overlapping dialogue, natural interactions, and interruptions.
- Improved Contextual Understanding: The ability to speak while correctly understanding stress, speaking rate, and sentence meaning is further enhanced.
Where and How Can It Be Used?
Eleven V3 is currently available through the company's website and mobile app. Users can access this technology by logging in to these platforms. However, it is not yet available as an API (Application Programming Interface), meaning developers or companies will have to wait a little while to directly integrate it into their systems.
AI Agent Interaction: 'Agent Transfer' Feature
ElevenLabs is constantly working on new technologies. In April, the company also launched a new enterprise-focused feature called Agent Transfer. This is part of the company's conversational AI system, where two AI agents can interact and transfer data to each other.
Through this feature, if one agent is not capable of providing specific information, it can transfer the conversation to an agent more proficient in that subject.
