Microsoft releases MAI Code 2 and expands its specialist AI lineup

Microsoft announced new coding, security, image and speech models at Ignite on November 17, led by MAI Code 2 and its smaller Flash variant. The company describes the broader launch as a family of seven new models, emphasizing lower operating costs for enterprise workloads.
MAI Code 2 and Code 2 Flash are available in GitHub Copilot. Code 2 Flash also has a limited-access preview in Microsoft Foundry. Microsoft positions Code 2 for debugging and changes across multiple files, while Flash targets routine edits and explanations. It says neither is yet tuned for fundamental science or very long-running agent tasks.
Across four coding benchmarks, Microsoft reports Code 2 at 98% of GPT-6.1 Sol’s performance for 40% of its cost. Those are company-reported comparisons, not a guarantee for customer workloads: Microsoft’s calculations use Code 2’s 50% promotional price, and results vary with configuration and pricing assumptions.
The rollout is staggered. MAI Cyber 2 Flash is scheduled to reach eligible customers through Microsoft security products, including Codename MDASH, starting in mid-December. It is not a general public release today.
Microsoft says MAI Image 2.7 and Image 2.7 Flash are available in Foundry. Its first live speech-to-speech model, MAI Realtime 1, starts at $0.01 per audio minute, with a demonstration and interest registration through MAI Labs.
For enterprise buyers, the practical distinction is between models they can start using now and announced capabilities still awaiting rollout. Microsoft’s pitch also shifts the comparison from benchmark scores alone to the cost of repeatedly running specialized tasks.
Sources and citations
Revision history
Published .