Microsoft AI, the tech giant’s research lab, has announced its launch three basic AI models; on Thursday that can generate text, voice and images.
The release marks Microsoft’s continued push to build its own multimodal AI model stack — and compete with rival AI labs — even as it remains tied to OpenAI.
MAI-Transcribe-1 transcribes speech in 25 different languages to text and is 2.5 times faster than Microsoft’s Azure Fast offering, according to a company press release. MAI-Voice-1 is a sound production model. This voice model allows users to create 60 seconds of audio in one second and allows users to create a custom voice. MAI-Image-2 is a video production model.
MAI-Figure-2 originally released on MAI Playgrounda new major language model testing software, on March 19. Now, all three models are live in the Microsoft Foundry, and the transcription and voice models are also available in the MAI Playground.
The models were developed by Microsoft’s MAI Superintelligence teaman AI research team led by Mustafa Suleyman, the CEO of Microsoft AI, created and announced in November 2025;
“At Microsoft AI, we build humanistic AI. We have a distinct point of view when building our AI models — putting people at the center, optimizing how people actually communicate, training for practical use,” Suleiman wrote in blog post. “You’ll see more models from us soon in Foundry and directly in Microsoft products and experiences.”
In an increasingly crowded LLM market, MAI hopes that a selling point for these models is that they are cheaper than Google’s and OpenAI’s, the company wrote in the blog post.
Techcrunch event
San Francisco, California
|
13-15 October 2026
MAI-Transcribe-1 starts at $0.36 per hour. MAI-Voice-1 starts at $22 per 1 million characters and MAI-Image-2 starts at $5 for 1 million characters for text input and $33 for 1 million characters for image output.
Despite releasing its own models, Suleyman confirmed Microsoft’s commitment to working with OpenAI on a interview with VentureBeat — even though a recent renegotiation of that partnership allowed Microsoft to actually pursue this super-espionage investigation, Suleiman told The Verge.
Microsoft has invested more than $13 billion in its AI research lab and hosts its models in its various products through a multi-year partnership. Microsoft takes the same stance with chips. produces its own and also buys from external players.
