Event ·Multi-source

11Labs发布语音识别模型Scribe v3与v3 Live

11Labs发布了两款语音识别模型:Scribe v3和Scribe v3 Live。v3是离线转写版本,v3 Live面向实时场景,主打低延迟流式识别。 从产品形态看,v3适合对准确率要求高的转写任务,如字幕生成、会议记录等;v3 Live则针对直播、语音Agent、实时翻译等需要边说边识别的场景。11Labs此前以语音合成产品ElevenLabs TTS著称,此次向语音识别方向扩展,补齐了语音链路中的一环。 推文作者认为语音识别领域目前领先的是11Labs和Gemini。对于需要搭建语音对话或实时转写服务的开发者来说,v3 Live与OpenAI Whisper、Deepgram等流式方案构成了新的对比选项。

Current read

11Labs这次一口气出了离线和实时两个语音识别模型,做语音Agent、字幕、会议记录的开发者可以直接对比试试,和Gemini、Whisper比谁的准确率更高。

11 sources52° AI heatLast updated 2026/11/11 01:54:00

Evidence

11 sources
Related source 1ElevenLabs
View source
Related source 2TestingCatalog
View source
Related source 4arXiv: OpenAI
View source
Related source 6Mustafa Suleyman
View source
Related source 7Artificial Analysis
View source
Related source 9kimmonismus
View source
Related source 10Anthropic
View source

Conflict check

Dedup data only shows that these sources discuss the same event, not that they agree. There is no structured supporting or refuting evidence yet, so no conflict is inferred automatically.