Eleven v3
lastmod 2025-06-09created 2025-06-091 edits
About
- Eleven v3 (Alpha) is the most expressive text-to-speech (TTS) model ever, with precise control down to emotions and voice effects
- Audio tags let you freely combine various voice elements such as emotion, tone of speech, direction, and sound effects
- Can generate natural audio of conversations between multiple speakers, and supports near-human voices in more than 70 languages
- Compared to v2, the range of vocal emotion and the scope of applicable effects have expanded greatly; UI users get an 80% discount until the end of June 2025
- API support is coming soon; various voice and situational tags can be found in the prompting guide