ElevenLabs 发布 v4 语音模型，并推出实时 Turbo 版本
原标题：ElevenLabs' new v4 speech model makes AI voices more expressive and consistent
AI 摘要
ElevenLabs 发布新一代语音模型 Eleven v4，能更准确地遵循语气、节奏等指令，并在长篇制作中保持音色一致，支持超过 90 种语言，单次请求最多处理 1 万字符。同时推出面向实时语音智能体的 Turbo 版本，约 150 毫秒开始发声，公司称快于 Cartesia Sonic 3.6 和 OpenAI GPT-4o mini TTS。两款模型已在 ElevenAgents、ElevenCreative 和 API 上线，并推出限时降价。
正文节选
ElevenLabs' new v4 speech model makes AI voices more expressive and consistent Key Points - Elevenlabs released Eleven v4, a speech model that follows direction cues more accurately and keeps voices consistent across long productions. - The new version supports more than 90 languages, improves voice cloning, and handles up to 10,000 characters per request. - Eleven v4 Turbo targets voice agents and responds in 150 milliseconds, which Elevenlabs says is faster than competing models. Elevenlabs is