Thursday, 19 November 2026 · Morning edition Wake Up to What Matters
Technology

Microsoft introduces affordable AI models at Ignite conference in San Francisco

At its Ignite conference in San Francisco on November 17, Microsoft unveiled new AI models, including MAI Code 2 and MAI Image 2.7, aimed at providing cost-effective solutions for coding and image editing, enhancing efficiency for enterprise customers.

Microsoft introduced several new artificial intelligence models at its Ignite conference in San Francisco on November 17, aimed at providing cost-effective alternatives to existing systems. The new models include MAI Code 2, MAI Code 2 Flash, MAI Cyber 2 Flash, and MAI Image 2.7, designed for various applications such as coding, cybersecurity, and image editing.

Microsoft launches MAI Code 2 and related models

MAI Code 2 is positioned as a competitive solution for coding tasks, achieving 98 percent of OpenAI’s GPT-6.1 Sol performance across four benchmarks at approximately 40 percent of the cost. It also reached 96 percent of Anthropic’s Claude Opus 5.5 performance while costing just 5 percent of that model. MAI Code 2 Flash is designed for simpler coding tasks, performing comparably to GPT-6 Luna at 65 percent of its cost and Claude Haiku 5.5 at 20 percent.

Both coding models are integrated into GitHub Copilot, with MAI Code 2 available through various subscription plans including Pro, Pro+, Max, Business, and Enterprise. MAI Code 2 Flash is accessible to Free and Student users and is currently in limited preview on Microsoft Foundry. Additionally, MAI Cyber 2 Flash reportedly matched GPT-6 Astra’s performance on the CyberGym benchmark for under 2 percent of the cost per task, although access remains restricted to a select group of customers.

MAI Image 2.7 shows improved performance

MAI Image 2.7, released as a preview on October 27, ranked second on the AA-Image-Editing v2.1 leaderboard with an Elo score of 1,193, just behind OpenAI’s GPT Image 2.5 Sunburst. This model demonstrated enhancements across all measured use cases, with significant improvements in composition, framing, and text-editing performance. MAI Image 2.7 is available through platforms such as MAI Playground and Arena.

Microsoft AI chief Mustafa Suleyman indicated that the company is also developing a general-purpose frontier model, although no launch date has been disclosed. He emphasized the company’s commitment to providing specialized models that can help reduce costs and improve efficiency for enterprise customers, stating, “This isn’t just one model. It’s not a flash in the pan.”

Background on Microsoft’s AI strategy

Microsoft’s approach reflects a shift towards offering specialized AI models rather than a single comprehensive solution. This strategy aims to meet the needs of businesses looking for effective AI solutions without incurring high operational costs. By bundling these models into existing products like GitHub and security tools, Microsoft seeks to maintain control over costs while enhancing the usability of its AI offerings.

As the market evolves, the introduction of these cost-effective models may pressure standalone AI vendors to adjust their pricing strategies, particularly as Microsoft positions itself as a competitive player in the AI landscape.