Microsoft's 3 new AI models priced below Google, OpenAI
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Microsoft AI released three foundational models for text, voice, and image generation on Thursday, positioning them as cheaper than offerings from Google and OpenAI, the company said in its blog post.
- MAI-Transcribe-1 transcribes speech across 25 languages at 2.5x the speed of Microsoft's Azure Fast transcription service, starting at $0.36 per hour.
- MAI-Voice-1 generates 60 seconds of audio in one second and supports custom voices, priced at $22 per 1 million characters.
- MAI-Image-2, described by Microsoft as a 'video-generating model' in its release, costs $5 per 1 million tokens for text input and $33 per 1 million tokens for image output on Microsoft Foundry.
- The models were built by Microsoft's MAI Superintelligence team, formed in November 2025 and led by CEO of Microsoft AI Mustafa Suleyman; all three are available on Microsoft Foundry, with the transcription and voice models also in MAI Playground.
- Mustafa Suleyman reaffirmed Microsoft's commitment to its OpenAI partnership in a VentureBeat interview, noting that a recent renegotiation of that deal allowed Microsoft to pursue its own superintelligence research; Microsoft has invested more than $13 billion in OpenAI.
Why it matters: A recent OpenAI partnership renegotiation gave Microsoft room to build competing in-house models, and the company is using that latitude to undercut both Google and OpenAI on price. Enterprise customers on Microsoft Foundry now get a multimodal AI stack starting at $0.36 per hour for transcription, reducing Microsoft's single-vendor exposure.
Ask SkimNews
