Our new, fastest model generates speech at ≈400ms latency and is over twice as fast as our V1 models. It also doesn't compromise on quality which stays on par with Multilingual V2.
For users of VoIP services, we now also support mulaw 8khz output with an even greater speed boost. See our API documentation to learn more.
We're working on adding multilingual support to Turbo v2, as well.
If you need help with integration or would like to speak about scale and support, feel free to contact our sales team.
Explore articles by the ElevenLabs team
Product
Eleven Music: new tools for exploring, editing and producing music with AI
Introducing a set of updates that expand what creators and developers can build with Eleven Music.
Agents Platform Stories
Building clinical-grade voice agents for Pharma
Increasing physician reach by 30% and cutting admin time by 10 hrs/week