US Edition
Microsoft Builds Its Own AI Model Stack To Reduce OpenAI Dependence
The models are priced to lower Microsoft’s AI costs and strengthen its position against OpenAI, Google and Anthropic.
- On Thursday, Microsoft launched three foundational AI models—MAI-Transcribe-1, MAI-Voice-1, and MAI-Image-2—built entirely in-house and available via Microsoft Foundry to compete directly with frontier labs.
- A contract renegotiation with OpenAI in September 2025 freed Microsoft to independently pursue artificial general intelligence, allowing the company to build proprietary models while retaining license rights to OpenAI's offerings through 2032.
- Microsoft AI CEO Mustafa Suleyman said MAI-Transcribe-1 achieves 3.8% Word Error Rate across 25 languages while using approximately 50 percent lower GPU cost than competitors, with the model built by fewer than 10 engineers.
- Pricing reflects a deliberate competitive strategy: MAI-Voice-1 costs $22 per 1 million characters and MAI-Image-2 costs $5 per 1 million tokens, undercutting Google and Amazon to capture enterprise market share.
- Suleyman plans to deliver world-class models across all modalities by next year, ensuring Microsoft achieves full AI self-sufficiency while continuing to host third-party offerings through its cloud platform.
14 Articles
14 Articles
East Bay Times
Australian Financial ReviewMicrosoft aims to create large cutting-edge AI models by 2027
Microsoft Corp. aims to develop large, cutting-edge artificial intelligence models by next year, part of a push to build in-house alternatives to the most powerful AI tools from OpenAI and Anthropic.
Microsoft Challenges OpenAI With Faster, In-House 'MAI' Models - Microsoft (NASDAQ:MSFT)
Microsoft (MSFT) unveiled its MAI-Transcribe-1 and MAI-Voice-1 models to boost AI speed and self-sufficiency. See why analysts maintain a "Buy" rating ahead of April earnings. Importance Rank: 1
Microsoft launches 3 new AI models in direct shot at OpenAI and Google
Microsoft on Wednesday launched three new foundational AI models it built entirely in-house — a state-of-the-art speech transcription system, a voice generation engine, and an upgraded image creator — marking the most concrete evidence yet that the $3 trillion software giant intends to compete directly with OpenAI, Google, and other frontier labs on model development, not just distribution.The trio of models — MAI-Transcribe-1, MAI-Voice-1, and …
Coverage Details
Bias Distribution
- 72% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium
