Transcribing speech is one thing. Knowing when someone is done speaking is another. 🎙 Voxtral 3’s ...
Mistral AI(@MistralAI)113 字 (约 1 分钟)
65
Voxtral 3的端到端语音识别技术提升了语音交互的响应能力,但技术细节披露不足。
入选理由:Voxtral 3新增端点检测功能可识别用户结束发言时刻
FeaturedTweet#语音识别#AI模型#Mistral AI英文