Microsoft MAI-Transcribe-2 hits near-top accuracy at $1.67 per 1,000 minutes
As seen on the 24/7 Wall St. homepage on September 3, 2026.
Microsoft is now selling near-best-in-class transcription at $1.67 per 1,000 minutes, squeezing the pricing power of standalone speech AI vendors.
Microsoft AI has released MAI-Transcribe-2, achieving 2.0% AA-WER (#2) at ~411x real time (#2) for $1.67 per 1,000 minutes, placing it on the accuracy-speed Pareto frontier at one of the lowest prices among high-accuracy models MAI-Transcribe-2 is @MicrosoftAI’s latest speech https://t.co/QRE68kWVjQ
- Replies6
- Reposts3
- Likes30
Continue ReadingShow less
MAI-Transcribe-2 lands at a 2.0% AA-WER score, placing it second on Artificial Analysis's accuracy rankings while also clocking in second for speed at roughly 411 times real time. Sitting at both extremes of the accuracy-speed Pareto frontier is rare, and Microsoft is doing it at one of the lowest prices among high-accuracy models.
At $1.67 per 1,000 minutes, Microsoft is undercutting the pricing power that standalone speech AI vendors have leaned on, making it harder for smaller competitors to justify a premium when a near-best-in-class alternative comes bundled with the broader Microsoft cloud ecosystem.
For investors in dedicated speech AI companies, the competitive pressure is direct and immediate. When a hyperscaler matches or exceeds specialist accuracy while pricing aggressively, the addressable market for those specialists contracts. The launch of MAI-Transcribe-2 is a signal that Microsoft is treating speech transcription as a commodity it intends to own.
The metric to track going forward is whether competing vendors respond with price cuts of their own or attempt to differentiate on domain-specific accuracy and features that a general-purpose model does not match. Either path narrows their margins.
Mentioned: MSFT