Microsoft AI Launches Three Models to Challenge OpenAI and Google

⚡

Key Takeaways

1Microsoft AI has unveiled three new AI models capable of generating text, voice, and images, strengthening its position in the market.

2The models MAI-Transcribe-1, MAI-Voice-1, and MAI-Image-2 are available on Microsoft Foundry and MAI Playground, offering enhanced performance.

3Mustafa Suleyman, CEO of Microsoft AI, claims that these models are designed to be more affordable than those from Google and OpenAI.

💡Why it matters — Microsoft is looking to diversify its AI offerings while maintaining its strategic partnership with OpenAI, thereby influencing the LLM market.

Microsoft AI Strengthens Its Position with Three New Models

Microsoft AI, the artificial intelligence research branch of Microsoft, recently announced the launch of three foundational AI models. These models are designed to generate text, voice, and images, marking a significant advancement in Microsoft's strategy to compete with other AI giants while remaining a partner of OpenAI.

Impressive Performance

The MAI-Transcribe-1 model stands out for its ability to transcribe speech into text in 25 different languages, with a speed 2.5 times faster than Microsoft's Azure Fast offering. MAI-Voice-1, on the other hand, can generate 60 seconds of audio in one second and offers the ability to create custom voices. Finally, MAI-Image-2 is a video generation model, initially launched on the testing software MAI Playground on March 19.

Availability and Accessibility

The three models are now accessible via Microsoft Foundry, with the transcription and voice models also available in MAI Playground. These tools were developed by the MAI Superintelligence team, led by Mustafa Suleyman, CEO of Microsoft AI, since their formation in November 2025.

Human-Centered AI

In a blog post, Suleyman emphasized Microsoft AI's humanistic approach, aiming to optimize human communication and provide practical solutions. He also announced the upcoming arrival of new models in Foundry and Microsoft products.

A Competitive Market

In a context where the LLM market is becoming increasingly saturated, Microsoft hopes to stand out by offering models that are more affordable than those from Google and OpenAI. Pricing starts at $0.36 per hour for MAI-Transcribe-1, $22 for 1 million characters for MAI-Voice-1, and $5 for 1 million tokens for text input and $33 for image output for MAI-Image-2.

Commitment to OpenAI

Despite developing its own models, Microsoft maintains its commitment to OpenAI. A recent renegotiation of their partnership has allowed Microsoft to continue its research in superintelligence while continuing to invest in joint projects. Microsoft has already invested over $13 billion in this partnership, integrating its models into various products while adopting a similar strategy for its chips, producing its own while sourcing from other suppliers.