ElevenLabs 推出 Eleven Realtime 语音智能体模型
原标题：Introducing Eleven Realtime: a simpler way to build natural voice agents
AI 摘要
ElevenLabs 发布 Eleven Realtime 系列语音到语音模型，包含面向低成本自然对话的 v1 Mini 和面向复杂任务的 v1 Max，支持 90 多种语言及 17000 多个音色。该模型基于端到端调优的级联架构，兼容 OpenAI Realtime API，开发者只需更换端点和模型名即可迁移。在 Artificial Analysis 的评测中，v1 Max 在 Big Bench Audio 上准确率 98.8%、平均 0.57 秒开始说话，并在准确率与速度的帕累托前沿上表现突出。
正文节选
Building a voice agent that sounds natural today requires a lot of decisions. You have to make choices on individual components like speech recognition, voice activity detection, turn-taking, the underlying LLM, and text to speech. Building this stack is a project in itself, and getting all the components to work well together takes more testing and tuning. Eleven Realtime is a new class of Speech to Speech models that removes those hurdles. Each model is a complete voice agent, optimized for se