Microsoft AI's in-house MAI model family covering reasoning, code, speech, transcription, image and cybersecurity.
Updated EU status checked Last shown
Sources: the vendor's own pages (linked above) and creator videos at the linked timestamps. EU status is information, not legal advice; see how we check.
MAI is a family of foundation models developed in-house by Microsoft AI. The website lists MAI Thinking 1 (reasoning, with reported top SWE-Bench Pro results at a mid-weight price), MAI Code 1.1 Flash (a lightweight agentic coding model built into GitHub Copilot and VS Code), MAI Transcribe 2 (domain-specific transcription), MAI Voice 2 (expressive, low-latency speech), MAI Image 2.6 (text and photo-prompted image generation) and MAI Cyber 1.1 Flash (a cybersecurity model inside the MDASH vulnerability identification and remediation harness).
Microsoft also offers Custom Models, which fine-tune an MAI model on a customer's data with stated privacy and control. Creator coverage describes the launch as seven models, including a flagship thinking model and a faster Transcribe release and a 15-language Voice release; the named versions on the site have since moved on, so version numbers may differ from early coverage. Benchmark claims (Artificial Analysis, Arena, FLEURS) come from Microsoft and have not been independently verified here.
The models are positioned for integration into Microsoft platforms such as GitHub Copilot and VS Code. Weights are not published, and the pages fetched did not give pricing or API documentation details.
Watch the moment: the exact second top creators show it. Videos load only when you press play, via youtube-nocookie.com.
Microsoft AI announced seven in-house models including a flagship thinking model (preferred over Sonnet 4.6), MAI Transcribe 1.5 (state-of-the-art accuracy, 5x faster) and MAI Voice 2 (15 languages). Suleyman says the goal is self-sufficiency and ethically licensed training data.
Closest neighbours by meaning, not just by category.
Model · Microsoft
Microsoft's coding-focused AI model in the MAI family, announced with few public details.
Model · Microsoft
Microsoft AI's mixture-of-experts reasoning model, in private preview via Microsoft Foundry.
Model · Microsoft
Microsoft's coding-focused language model, positioned as a faster, more token-efficient alternative to Claude Haiku 4.5.
Model · Microsoft
Microsoft AI's text-to-speech voice model with expressive output across multiple languages.
Model · Microsoft
Microsoft AI's low-latency text-to-speech model in the MAI model family, designed for long-form expressive speech.
Model · Microsoft
Microsoft AI's pre-trained foundation model with about 80B active parameters; post-training unfinished and not yet available.
