Microsoft Launches Cheaper, Faster Speech Recognition
π‘Microsoft's new model claims 60-language coverage and $0.10-per-hour transcription for messy enterprise audio.
β‘ 30-Second TL;DR
What Changed
The early-bird price is $0.10 per audio hour, down roughly 72% from MAI-Transcribe-1's $0.36 per hour.
Why It Matters
The sharp price reduction could lower the cost of call-center analytics, meeting transcription, compliance workflows, and captioning at scale. It also strengthens Microsoft's strategy of replacing OpenAI-powered components with internally developed models across its products.
What To Do Next
Run a representative call-center audio sample through MAI-Transcribe-2 on Microsoft Foundry and compare cost, word error rate, diarization, and code-switching accuracy with your current provider.
Key Points
- β’The early-bird price is $0.10 per audio hour, down roughly 72% from MAI-Transcribe-1's $0.36 per hour.
- β’Language coverage increased to 60 languages from 43 in MAI-Transcribe-1.5 and 25 in the original release.
- β’Enterprise features include speaker diarization, word-level timestamps, keyword biasing, and automatic language identification.
- β’Verbatim and clean output modes support compliance transcripts as well as readable captions and notes.
- β’Code switching is designed for mixed-language conversations such as Hinglish and Spanglish.
Weekly AI Recap
Read this week's curated digest of top AI events β
πRelated Updates
AI-curated news aggregator. All content rights belong to original publishers.
Original source: VentureBeat β
This is a summary, not the original. Read the source, or get the weekly briefing.
Weekly AI briefing
One email a week. Unsubscribe anytime.