{"slug": "mai-code-2-and-code-2-flash", "title": "MAI Code 2 and Code 2 Flash", "summary": "Microsoft released MAI Code 2 and MAI Code 2 Flash, two coding models the company says deliver 98% of GPT-6.1 Sol's performance at 40% of the cost and 96% of Claude Opus 5.5's performance at just 5% of the cost, averaged over four benchmarks (DeepSWE 1.1, Terminal Bench 2.1, SWE-Atlas and FrontierCode v1.1). MAI Code 2 Flash, aimed at high-volume routine coding, matches GPT-6 Luna and Claude Haiku 5.5 at 65% and 20% of their cost across three benchmarks, and in blind frontend app evaluations developers preferred MAI Code 2's results over GPT-6 Sol 64% of the time. Both models are available in GitHub Copilot, with Code 2 Flash in limited-access preview in Microsoft Foundry and planned releases on OpenRouter, Vercel, Hermes, OpenCode and Kilo; Microsoft said the models are not yet tuned for fundamental science research or very long-horizon agentic work.", "body_md": "### MAI Cyber 2 Flash: Leading security. Lower cost\n\nMAI Code 2 and MAI Code 2 Flash: two new coding models built for everyday enterprise coding tasks, delivering leading performance at a fraction of the cost of competing models.\n\nMAI Code 2 is at the frontier of price-performance for common enterprise coding tasks and excels in taking code to production, deep debugging, multi-file implementation, repo-wide reasoning, and UI. It does all of this while also delivering changes code maintainers would merge, at a fraction of frontier cost.\n\nMeanwhile, MAI Code 2 Flash is designed for high-volume work such as code explanations, targeted edits, routine fixes, and rapid iterations, while bringing quality at a much lower cost.\n\nTo deliver the best price-performance, we’ve built these models to prioritize the coding work our customers do every day. They’re explicitly not yet tuned for harder problems, like fundamental science research and very long-horizon agentic work.\n\nAveraged over four widely used benchmarks that target the everyday enterprise coding tasks we see in GitHub Copilot (DeepSWE 1.1, Terminal Bench 2.1, SWE-Atlas and FrontierCode v1.1), MAI Code 2 gives you **98%** of the performance of GPT-6.1 Sol at **40%** of the cost, and **96%** of Claude Opus 5.5’s performance at just **5%** of the cost.[1](#fn-1)\n\nMAI Code 2 is also tuned for producing great dashboards and visualizations, and excels at visual coding. In blind frontend app evaluations, developers preferred its results over GPT-6 Sol **64%** of the time.[2](#fn-2)\n\nIn addition, MAI Code 2 Flash offers comparable performance to GPT-6 Luna and Claude Haiku 5.5 at **6****5%** and **20%** of the cost, across three coding benchmarks measuring day-to-day, high-volume routine coding tasks.[3](#fn-3)\n\nMAI Code 2 is also becoming the workhorse behind Microsoft products. From Code 2 Flash, we further optimized a single model, MAI Work Flash, to serve multiple Microsoft products across Microsoft 365 and FabricIQ. \n\n \n\n**Excel:** Our model is on par for accuracy and token efficiency with GPT-6 Luna across key internal Excel customer-scenario benchmarks.  \n\n**PowerPoint:** On key production benchmark scenarios that measure frequent PowerPoint tasks, our model performs on par with GPT-6 Luna in both quality and token efficiency.  \n\n**Word:** Our model performs similarly to GPT-6.1 Sol (while being significantly smaller) on 5 key internal benchmarks using ~42% fewer input tokens per call than GPT-6.1 Sol in production. \n\n**Teams:** Our model beats GPT-6 Luna at tracking decisions, filling knowledge gaps, and adapting to custom instructions using up to 28% fewer tokens on instruction following and tool calling. \n\n**FabricIQ:** In offline evaluation runs, our MAI model outperforms GPT-5.6 Sol, GPT-6 Astra, and Opus-5.5 at DAX query generation, the core capability behind “Chat with your Data” for Power BI semantic models. \n\nGet started\n\nMAI Code 2 and Code 2 Flash are available in GitHub Copilot, and Code 2 Flash is in limited-access preview in [Microsoft Foundry](https://azure.microsoft.com/en-us/products/ai-foundry). Soon, we’ll be releasing them on OpenRouter, Vercel, Hermes, OpenCode, and Kilo. \n\nLearn more about customizing MAI Code 2 on our [Custom Models page.](https://microsoft.ai/models/microsoft-frontier-tuning/) \n\n**About these results**\n\n*Results reflect evaluations on specific coding tasks and configurations. Performance and cost comparisons vary by workload, model configuration, reasoning effort, and pricing assumptions.*", "url": "https://wpnews.pro/news/mai-code-2-and-code-2-flash", "canonical_source": "https://microsoft.ai/news/introducing-mai-code-2-and-mai-code-2-flash/", "published_at": "2026-11-17 18:25:13+00:00", "updated_at": "2026-11-18 18:12:43.018451+00:00", "lang": "en", "topics": ["ai-products", "large-language-models", "developer-tools", "ai-tools"], "entities": ["Microsoft", "MAI Code 2", "MAI Code 2 Flash", "MAI Work Flash", "GitHub Copilot", "Microsoft Foundry", "GPT-6.1 Sol", "Claude Opus 5.5"], "also_reported_by": [], "takeaway": "MAI Code 2 delivers 98% of GPT-6.1 Sol's benchmark performance at 40% of the cost and 96% of Claude Opus 5.5's at 5% of the cost, but is explicitly not tuned for very long-horizon agentic work or fundamental research.", "recommendation": null, "alternates": {"html": "https://wpnews.pro/news/mai-code-2-and-code-2-flash", "markdown": "https://wpnews.pro/news/mai-code-2-and-code-2-flash.md", "text": "https://wpnews.pro/news/mai-code-2-and-code-2-flash.txt", "jsonld": "https://wpnews.pro/news/mai-code-2-and-code-2-flash.jsonld"}}