ElevenLabs Eleven Realtime: speech-to-speech models for voice agents
ElevenLabs launched Eleven Realtime v1 Mini and v1 Max — single speech-to-speech models that fold recognition, turn-taking, the language model, and voice into one system for building voice agents.
Why it matters
One model replaces the STT+LLM+TTS pipeline: Max scores 98.8% on Big Bench Audio with 0.57s time-to-first-audio, and an OpenAI-compatible Realtime API makes migration nearly code-free.
What you could build with it
A small team building phone-based customer support could build its voice agent on Eleven Realtime v1 Max instead of wiring speech recognition, a language model, and speech synthesis together — the single model handles interruptions and turn-taking natively, and the OpenAI-compatible Realtime API lets an existing integration migrate with minimal code changes.
Does it hold up?
Too early to judge for production: Artificial Analysis independently benchmarked the models (Max 98.8% Big Bench Audio, 998 Speech Agent Arena), but no third-party production deployments have been reported yet and pricing is per-minute plus tokens.
Built with ElevenLabs Eleven Realtime
- ElevenLabs releases two new speech-to-speech models for real-time AI conversationsarticle · Launch coverage with benchmark context: Max at 98.8% Big Bench Audio and 0.57s response time, Mini tuned for natural conversation flow and lower cost.
- ElevenLabs brings faster voice agents with Mini and Maxarticle · Breakdown of the two-model lineup: Mini leads on Full-Duplex-Bench (95.8% vs 94.1%) for conversational dynamics, Max dominates reasoning and tool-use tasks.
- ElevenLabs launches Eleven Realtime: voice agents from a single modelarticle · Analysis of the competitive angle: single-model architecture vs Mistral's Voxtral 3 three-component stack, with pricing per minute plus token costs.
Learn more
First spotted on article: source.
More AI models
OpenAI publishes 722 AI-generated math manuscripts from an unreleased modelOpenAI released 722 mathematical manuscripts in 372 result families to GitHub under Apache-2.0, produced by an…models · JEV 0.74Tencent Hunyuan Hy3: open 295B MoE with 21B active under Apache 2.0Tencent open-sourced Hy3, a 295B-parameter mixture-of-experts model with only 21B parameters active per token, under…models · JEV 0.73NVIDIA Isaac Lab 3.0 GA + Isaac OS open runtime + Agile ONE SN humanoidMajor Isaac platform expansion at GTC Berlin Oct 21: Isaac Lab 3.0 is generally available as the open-source robot…models · JEV 0.73StepFun Step-5-Preview — open BF16 weights releasedStepFun published the full BF16 weights for its flagship Step-5-Preview agentic model (600B-parameter sparse MoE, 27B…models · JEV 0.73NVIDIA Nemotron 3.5 Retriever collection: Embed 8B/1B, Rerank 8B, Parse 631MNVIDIA's Oct 21 collection drop for agentic retrieval: Nemotron-3.5-Embed in 8B BF16, 1B BF16 and 1B NVFP4…models · JEV 0.72Kimi K3.1 Pro: Moonshot's 2.8T open-weight flagship officially launchesMoonshot AI officially launched Kimi K3.1 Pro on Oct 23, a mixture-of-experts model with 2.8 trillion total / 104…models · JEV 0.72
Get the week's best AI launches, plus 3 ideas worth building
One email every Saturday. Ranked by traction, not hype. Free.