Fish Audio 的 S2.1 Pro 在日本正式上线，做日语语音代理延迟不到 150ms，还能日英混说，现在限时免费试。
Fish Audio 宣布在日本上线，其语音模型 S2.1 Pro 已出现在东京出租车场景中。该模型针对日语难点优化，包括音高重音、敬语和日英混说，支持客服、销售、日程安排等智能体，延迟低于 150ms。目前已覆盖 80 多种语言，拥有 800 万用户，提供云 API 和本地部署两种方式。S2.1 Pro 限时免费试用。
Fish Audio is live in Japan! 🇯🇵 Already heard us in the back of a Tokyo taxi? That’s S2.1 Pro. Ja...
Fish Audio is live in Japan! 🇯🇵 Already heard us in the back of a Tokyo taxi? That’s S2.1 Pro. Japanese is one of the hardest tests in voice AI. You can pronounce every word correctly and still sound wrong: pitch accent, prosody, keigo, jp ↔ en switching mid-sentence. → voice agents for support, sales, scheduling at sub-150ms → AI companions with actual personalities → game characters: cloning, dynamic dialogue, localization → dubbing / animation / audiobooks 8M+ users. 80+ languages. Cloud API or on-prem. Build with S2.1 Pro free, limited time 👇 Your browser does not support the video tag. 🔗 View on Twitter 💬 5 🔄 0 ❤️ 11 👀 778 📊 7 ⚡