Passing the Turing test for voice agents. We just shipped one of our biggest ElevenAgents update yet.
A set of powerful updates:
- Expressive mode with v3 conversational model: ultra-low latency Text-to-Speech optimised for real-time conversations - with finer control over delivery so agents can use a reassuring tone when a user sounds worried, or adjust cadence mid-conversation
- Optimized conversational model with state-of-the-art turn-taking: Scribe v2 real-time now captures intent and emotional cues more reliably - agents respond with better timing, fewer interruptions, more natural flow
- Available in 70+ languages - agents now speak a wider set of languages and dialect all in real-time
The upgrades amplify each other. Better transcription means agents understand context and emotion. Better Text-to-Speech means they respond with the right tone at the right moment.
Live today!