ElevenLabs 发布 Eleven Realtime 两款语音到语音模型，面向实时 AI 对话
Elevenlabs releases two new speech-to-speech models designed for real-time AI conversations
- 1ElevenLabs 发布两款语音到语音模型 Eleven Realtime，面向实时 AI 对话场景
- 2官方称 v1 Max 在 Big Bench Audio 基准得 98.8%，响应时间 0.57 秒，暂无独立验证
- 3v1 Mini 面向低成本自然对话，能分辨对方说完还是在说 mm-hmm；v1 Max 定位客服等复杂任务
- 4两款模型经 ElevenAgents 平台和 OpenAI 兼容 Realtime API 提供，按分钟加 token 计费
ElevenLabs 把语音到语音实时对话做成专门模型，实时语音智能体在延迟与准确率之间多了一个供应商选项。
- 影响：语音客服和实时语音智能体场景多了一个主打高准确低延迟的可选模型；OpenAI 兼容接口让替换现有方案的接入成本较低；按分钟加 token 计费与按次计费结构不同，需要重算单通对话成本。
- 参考：可考虑选一个客服通话场景做小规模试点，对比现有方案的延迟和应答准确率；值得按分钟加 token 的计费口径先算出单通电话成本，再决定是否切换。
- 依据：98.8% 基准得分和 0.57 秒响应均为 ElevenLabs 官方口径，暂无独立验证；排行榜对比来自官方引用的 Artificial Analysis 数据，榜单截图由官方提供。
Oct 20, 2026
Elevenlabs released two new speech-to-speech models called Eleven Realtime. The Max variant scores 98.8 percent on the Big Bench Audio benchmark with a 0.57-second response time. Elevenlabs says no model on the Artificial Analysis leaderboard is both faster and more accurate right now.
The smaller v1 Mini is built for natural conversations at lower cost and can tell when someone's done talking or just saying "mm-hmm." V1 Max targets more complex tasks like customer support.
Both models are available through the ElevenAgents platform and an OpenAI-compatible Realtime API. Pricing is per minute plus token costs, with v1 Max costing more than v1 Mini. Full details are on the Elevenlabs pricing page.
AI News Without the Hype – Curated by Humans
Subscribe to THE DECODER for ad-free reading, a weekly AI newsletter, our exclusive "AI Radar" frontier report six times a year, full archive access, and access to our comment section.
来源：The Decoder：AI News · the-decoder.com