Microsoft MAI Code 2
3 steps2 pieces3 Sources
Microsoft has launched its in-house AI coding models, MAI Code 2 and MAI Code 2 Flash, at Ignite 2026. The company claims these models deliver 96% of the performance of Anthropic's Claude Opus 5.5 at just 5% of the cost, aiming to reduce reliance on external providers like OpenAI and Anthropic. While Microsoft positions this as a move toward vertical integration and lower operational costs for GitHub Copilot and Azure, independent analysts caution that the benchmarks are self-reported and promotional pricing may not reflect long-term savings. The launch also introduces "Microsoft IQ" and redefines Copilot as an "operating system for work," integrating coding, security, and business processes into a single platform.
How it got here
Nov 17, 2026: Microsoft announces MAI Code 2 and Code 2 Flash at Ignite 2026, claiming near-frontier performance at significantly lower costs compared to Claude Opus 5.5 and GPT-6.1 Sol.
zdnet.com ↗Nov 17, 2026: GitHub releases MAI Code 2 and Code 2 Flash as generally available models in GitHub Copilot for Pro, Business, and Enterprise users.
github.blog ↗Nov 18, 2026: Industry reports highlight Microsoft's strategy to reduce dependency on OpenAI and Anthropic by using proprietary models for high-volume, routine coding tasks.
infoworld.com ↗