Microsoft takes on AI rivals with three new foundational models

In a bold move to enhance its AI capabilities, Microsoft has unveiled three new foundational AI models that generate text, voice, and images. This launch underscores the tech giant's commitment to developing a comprehensive suite of multimodal AI technologies, as it continues to navigate the competitive landscape alongside other AI leaders, while maintaining its partnership with OpenAI. The first model, MAI-Transcribe-1, revolutionizes speech transcription by converting spoken language into text in 25 different languages, boasting a speed that is 2.5 times faster than Microsoft's existing Azure Fast service. The second innovation, MAI-Voice-1, is designed for audio generation, allowing users to produce a minute of audio in just one second, and offers the ability to create custom voices. Completing the trio is MAI-Image-2, a model capable of generating video content. Initially introduced on March 19 through MAI Playground, a new platform for testing large language models, all three models are now accessible via Microsoft Foundry, with the transcription and voice capabilities also available in MAI Playground. These advancements are the result of efforts by Microsoft’s MAI Superintelligence team, led by Mustafa Suleyman, the CEO of Microsoft AI. The team was established in November 2025, with a focus on developing AI that prioritizes human-centered design. Suleyman highlighted this vision in a recent blog post, emphasizing the importance of crafting AI that aligns with natural human communication and practical application. As the landscape of large language models grows increasingly competitive, Microsoft aims to set itself apart by offering more affordable solutions compared to offerings from Google and OpenAI. Pricing for MAI-Transcribe-1 begins at $0.36 per hour, while MAI-Voice-1 is priced at $22 per million characters. MAI-Image-2 starts at $5 for one million tokens of text input and $33 for image output. Despite the introduction of these models, Suleyman reaffirmed Microsoft's dedication to its ongoing partnership with OpenAI, which has seen the company invest over $13 billion into the AI research lab. This collaboration allows Microsoft to integrate OpenAI's models into its product offerings while also developing its proprietary AI technologies, balancing in-house production with external sourcing of components.