News
Microsoft Brings MAI Code 2 and Code 2 Flash to GitHub Copilot, Teases Cyber 2 Flash at Ignite

At Ignite 2026, Microsoft unveiled its own MAI models for coding and cybersecurity. Code 2 and Code 2 Flash are already available in GitHub Copilot, with Microsoft saying Code 2 runs at a fraction of frontier-model cost and Code 2 Flash delivers near-flagship quality at a much lower cost.
Contents
Microsoft is expanding its in-house MAI model family. At Ignite on November 17, 2026, the company showed Code 2, its flagship model for agentic coding, a cheaper variant called Code 2 Flash, and Cyber 2 Flash, a model for security tasks. Both coding models landed in GitHub Copilot right away.
What Microsoft announced
According to the official Microsoft blog, Code 2 is a new flagship family of models for agentic coding. The company describes it as a model for everyday enterprise use, strong at taking code to production, deep debugging, multifile implementation, repo-wide reasoning and UI work. Microsoft says it does this at a fraction of frontier cost.
Code 2 Flash is meant for fast, high-volume work: code explanations, targeted edits, routine fixes and rapid iterations. The company says it delivers near-flagship quality at a much lower cost. These are the vendor's claims; the materials we read include no independent measurements or specific prices.
We are also innovating at the frontier with our own MAI model family that is both compute-efficient when it comes to pre-training and token-efficient when it comes to product deployment. - Satya Nadella, CEO of Microsoft
Availability in Copilot
A GitHub changelog entry says Code 2 is available in the Pro, Pro+, Max, Business and Enterprise plans. Code 2 Flash also comes to the Free and Student plans. GitHub notes that both models are available to eligible subscribers on annual plans. The models appear in the model picker in VS Code, Visual Studio, Copilot CLI, the cloud agent, the Copilot app, on github.com, in GitHub Mobile, JetBrains, Xcode and Eclipse.
Both models fall under usage-based billing. GitHub points to its Copilot models and pricing documentation for details, and the post itself includes no multipliers or rates. Administrators of Business and Enterprise plans manage access through the model policy in Copilot settings, and new models are enabled by default unless an administrator has turned off the global default or disabled a specific model.
Cyber 2 Flash and licensed weights
Microsoft calls Cyber 2 Flash its most advanced security model yet. It is a Flash-class model whose results, the company says, are ahead of frontier models at a fraction of the cost per task. It is designed together with the tools where it will run and is being integrated with Perception.
Nadella also announced MAI Licensed Weights. Trusted customers are to get access to Microsoft AI model weights. The license terms were not described in the text we read.
In the same keynote, Nadella mentioned the general availability of Microsoft-Decision-1.1, a model with multimodal support that Microsoft uses internally for routing, data labeling, personalization and content filtering, among other things. The company says more than 4,000 customers have used it.
What it means for dev teams
For companies using Copilot, this is another choice alongside models from outside providers. The real cost depends on the rates in GitHub's documentation, and quality has to be tested on your own code. Teams in Poland that pay for Copilot on a usage basis can compare Code 2 Flash with the cheaper models they use now for routine tasks. For the moment, the quality and cost figures come from Microsoft.
Strategically, the announcement shows that Microsoft wants its own models in Copilot alongside those from partners. Whether that translates into real savings will be shown by independent tests and public price lists.


