MAI-Transcribe-2
Top-3 accuracy at one of the lowest per-minute prices.
MAI-Transcribe-2 is Microsoft AI’s in-house speech-to-text model, combining near-best word error rate with very low pricing.
- Provider
- Microsoft AI
- Type
- Speech-to-text
- Released
- Sep 3, 2026
- Price
- ≈$1.67 per 1,000 minutes
- Input
- audio
- Output
- text
- Open weights
- No
Best for
- Extraction & classification
- High volume / low cost
- Enterprise
Strengths
- ≈2.0% WER
- ≈$1.67 per 1,000 minutes
More from Microsoft AI
- MAI-Image-2.6 — Microsoft AI. Microsoft AI’s in-house image model.