FishAudio 发布 Live 3：支持全双工语音的实时对话模型
FishAudio 这个 Live 3 能边听边说，不用像对讲机那样一人一句，语音智能体的打断体验会自然很多。vLLM 团队还主动喊话说开源后第一时间支持。做语音 Agent 的开发者可以先盯住它的开源动向。
FishAudio 推出 Live 3 模型，主打全双工语音交互：用户说话时模型可以同时聆听和回应，不必像对讲机那样轮流发言，首个版本即面向生产环境交付。
vLLM 项目公开祝贺，并表示期待该模型未来开源，愿以 vLLM-Omni 从第一天起提供推理支持。全双工能力解决了语音智能体常见的轮次打断问题，让对话更接近自然人声交互。
目前 FishAudio 尚未公布开源时间表，社区只能先等官方释放权重或 API 细节。若 Live 3 后续开源并接入 vLLM-Omni，开发者将能在自有服务器上部署低延迟全双工语音智能体，与依赖云端托管的方案形成差异化。
No more walkie-talkie voice agents 📻
@FishAudio just shipped Live 3: a full-duplex model that listens and speaks at the same time, production-ready on day one.
Congrats team! Hoping to see it open-sourced one day, and vLLM-Omni would be excited to support it from day 0 🚀
This report is categorized under AITOP's "Models" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows and US-China foundation model benchmarks and commercialization comparison. Key entities and technical topics under continuous tracking include FishAudio, Live 3, vLLM, 语音交互.
Source attribution: curated from vLLM (@vllm_project). Cross-source corroboration: 4 independent sources have covered this development (including marktechpost, Fish Audio, TestingCatalog), confirming industry-wide signal strength.
