Live drop
MAI-Transcribe-2-Streaming
Microsoft AIDropped Oct 1, 2026
Real-time streaming speech-to-text with first partials in ~100ms and top accuracy across 60 languages.
About this release
Live drop
Microsoft AIDropped Oct 1, 2026
Real-time streaming speech-to-text with first partials in ~100ms and top accuracy across 60 languages.
About this release
MAI-Transcribe-2-Streaming is Microsoft AI's first dedicated streaming speech-to-text model. It processes continuous audio streams, returning incremental partial transcripts that update in real time before committing final results, with automatic language detection and support for 60 languages. It ranks #1 on Artificial Analysis for both final and partial transcript accuracy while maintaining low latency suitable for voice agents, live captions, and dictation.
Read the official Microsoft AI launch pageFirst builds incoming — this lane fills in automatically as people start building with MAI-Transcribe-2-Streaming. Know one? Submit the URL.