AI BriefingElevenLabs launches Eleven Realtime v1 Mini and Max speech-to-speech models
·2026.10.21 09:00
Key point
The new models combine speech recognition, turn-taking, LLM, and voice into a single end-to-end system available via ElevenAgents and an OpenAI-compatible API.
Details
- ElevenLabs introduces Eleven Realtime, a new class of Speech to Speech models available in two variants: v1 Mini for natural conversations and v1 Max for complex tasks requiring reasoning and tool calls.
- v1 Max reports a 98.8% score on Big Bench Audio and an 87.1% task success rate, with a time to first audio of 570ms; v1 Mini reports a 95.8% score on Full-Duplex-Bench and a Speech Agent Arena Elo of 988, with a time to first audio of 620ms.
- The Realtime API is compatible with OpenAI's Realtime API, allowing developers to switch endpoints and model names while retaining existing event handlers and SDKs; it connects via WebSocket and does not support WebRTC at launch.
- Available in over 90 languages and accents, the platform supports enterprise compliance standards including SOC 2 Type II, ISO 27001, HIPAA, and GDPR, with options for EU Data Residency and Zero Retention modes.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.