← November 19, 2026 briefing

Microsoft AI releases seven MAI models for coding, security, image, voice, and decisions; claims MAI Code 2 hits 98% of GPT-6.1 Sol at ~40% of the price (community report)

Verified (fact)

According to a GeekNews summary, Microsoft AI released seven new models at once, spanning security, coding, image, voice, and decision-making. They are reportedly priced so that agents doing real engineering work can run 24 hours a day affordably. The lineup includes MAI Code 2, MAI Cyber 2 Flash for continuous security monitoring, MAI Image 2.7 for image editing, and Microsoft's first real-time speech-to-speech model. MAI Code 2's arrival in Copilot was covered earlier; this time the remaining models and comparison figures were newly disclosed.

Needs fact-check (claim)

Microsoft claimed that on the average of its own enterprise coding benchmarks, MAI Code 2 delivers 98% of GPT-6.1 Sol's performance at about 40% of the price. It said MAI Cyber 2 Flash achieves higher quality than GPT-6 Astra on CyberGym-E2E at less than 2% of the per-task cost, and that MAI Image 2.7 ranks first in Artificial Analysis's image editing evaluation.

Why it matters (interpretation)

Microsoft has begun promoting its own domain-specific models on price-performance rather than relying on OpenAI models. Competition over the operating cost of always-on agents is heating up.

Confidence medium · official source pending

Sources

More about Microsoft