Microsoft upgrades medical-patient dialogue AI transcription engine
The story so far
Microsoft has upgraded its medical-patient dialogue AI transcription engine MAI-Transcribe-2, reducing multilingual error rates by 50%, now handling 100% of medical-patient dialogue transcription tasks in Dragon Copilot. The model can identify multiple languages and common mixed-language scenarios such as Chinese-English and Hindi-English, and maintains high recognition quality in noisy environments. MAI-Transcribe-2 supports speaker differentiation, word-level timestamps, keyword bias, and configurable transcription styles, and supports two output modes: 'word-by-word' or 'concise', suitable for different needs such as compliance archiving, analysis, and medical record writing.
AI-written from the reports · Updated 22 hours ago
Coverage timeline
Follow the coverage to see every side of the story.
- ITHome · AIMicrosoft has upgraded its medical-patient dialogue AI transcription engine MAI-Transcribe-2, with multilingual error rates dropping directly by 50%.
Microsoft announced at Ignite 2026 that it is adopting its self-developed speech recognition model MAI-Transcribe-2 to upgrade Dragon Copilot Transcribe. The model achieves a leading performance with an average word error rate of 5.2% across 60 languages in the FLEURS benchmark, and its long audio processing speed is up to 10 times faster than some competitors. It now handles 100% of the doctor-patient conversation transcription tasks in Dragon Copilot.
Heat over time
- Now
- 5
- Peak
- 1018 Nov, 08:00
- 24h change
- –
Hover or tap the chart to see each hour; use the arrow keys on a keyboard.