Microsoft Takes on OpenAI and Anthropic With Cheaper AI Coding Models at Ignite 2026

Microsoft Takes on OpenAI and Anthropic With Cheaper AI Coding Models at Ignite 2026

Microsoft introduced two proprietary AI coding models, MAI Code 2 and MAI Code 2 Flash, at its Ignite 2026 conference in San Francisco on November 17, alongside a new cybersecurity model called MAI Cyber 2 Flash. Both coding models are now generally available through GitHub Copilot, while the security model will initially reach a limited group of customers. The announcements were part of a broader update covering Microsoft's AI models, Copilot platform, enterprise agents and cloud infrastructure.

The new coding models are designed to give businesses alternatives for software development tasks with different performance and cost requirements. MAI Code 2 targets more demanding enterprise coding work, while Code 2 Flash prioritizes efficiency for routine programming activities and higher-volume workloads.

During his opening keynote, Microsoft chairman and CEO Satya Nadella described Code 2 as operating "at the frontier of cost-performance for everyday enterprise coding tasks."

According to Microsoft, MAI Code 2 achieves 98% of GPT-6.1 Sol's coding performance at approximately 40% of the cost. The company also claims it reaches 96% of Claude Opus 5.5's performance at roughly 5% of the price, based on results averaged across DeepSWE 1.1, Terminal Bench 2.1, SWE-Atlas and FrontierCode v1.1.

Code 2 is optimized for common enterprise programming assignments and frontend development. Microsoft acknowledged that the model has not yet been specifically optimized for scientific research or particularly long-running agent workflows.

MAI Code 2 Flash is intended for less demanding tasks, including straightforward code modifications and fixes. Microsoft says it matches GPT-6 Luna's performance at 65% of the cost and Claude Haiku 5.5 at 20% of the cost.

Both models can be selected in GitHub Copilot and used through Copilot Fusion. Code 2 Flash is also available in limited preview through Microsoft Foundry. Microsoft did not provide pricing details for the coding models in its keynote.

The company also expanded its security-focused AI offerings with MAI Cyber 2 Flash, which Nadella described as "our most capable security model yet."

Microsoft reported that Cyber 2 Flash scored 92.3% on CyberGym-E2E, compared with 91.5% for GPT-6 Astra, while completing tasks at less than 2% of the cost. The model supports a context window of one million tokens and is priced at $0.10 per million input tokens and $0.50 per million output tokens.

Unlike the coding models, Cyber 2 Flash is being introduced through Microsoft's security products rather than as a general-purpose offering. The company plans to deploy it to eligible customers through Codename MDASH beginning in mid-December, replacing the existing model used by that service.

Microsoft is also integrating the new security capabilities with Perception, which has reached general availability.

Beyond coding and cybersecurity, Microsoft announced additional models covering image generation, voice interaction and AI request routing.

MAI Image 2.7 achieved the top position on Artificial Analysis's image-editing evaluation, according to the company, surpassing GPT Image 2.6 Sunburst at approximately one-fifth of the cost. A lighter version, MAI Image 2.7 Flash, is priced at about $10 per 1,000 images, compared with approximately $35 for the full model.

Both image models are available in public preview through Foundry and partners including fal, Genspark, OpenRouter and Vercel. The Flash version also powers a new Canvas workspace within MAI Labs.

For voice applications, Microsoft introduced MAI Realtime 1, its first model designed for direct speech-to-speech interaction. Pricing begins at $0.01 per minute of audio, with interested users able to register through MAI Labs.

The company also highlighted Dragon Copilot Transcribe, which uses MAI Transcribe 2 and currently processes 10 million patient encounters each month.

Microsoft-Decision-1.1, a model designed to route requests to larger AI systems only when necessary, is now generally available for enterprise use. It received a score of 68 on Artificial Analysis's Decision Index, compared with 62 for the previous version.

Microsoft is expanding access to models developed by other companies as well. Nadella said OpenAI's GPT-6.2 models and Meta's latest models are coming to Microsoft Foundry, giving enterprise customers additional options alongside Microsoft's own AI offerings.

A new initiative called MAI Licensed Weights will allow selected enterprise customers to customize Microsoft's model weights using proprietary data and deploy those models within their own Azure environments.

The announcements also included significant changes to Microsoft's Copilot experience. Nadella outlined a structure built around Copilot Home, Code, Autopilot and Work IQ, bringing together conversational assistance, programming tools and enterprise workflows.

Microsoft is adding integrations with business applications, beginning with Dynamics 365, and is bringing Salesforce capabilities into Copilot. The company also introduced Copilot Workspaces, a collaborative environment for working with AI, although it did not specify when the feature would become available.

"Copilot is the new OS for work," Nadella said during the keynote.

Microsoft IQ is receiving hundreds of additional connectors as the company brings together information from Microsoft Fabric, Foundry and Microsoft 365.

Enterprise agent management was another major focus. Nadella said Agent 365 already manages more than 50 million agents, reflecting the scale of Microsoft's existing AI agent environment.

The company introduced Agent 365 Business Analytics to help organizations monitor and manage their use of AI across business operations.

"You can think of it as your ERP system for AI usage," Nadella said.

Microsoft also announced the AI Transparency Initiative, which includes collaboration with the National Institute of Standards and Technology's Agent Standards Initiative and 40 organizations.

On the infrastructure side, Microsoft announced general availability of its next-generation Cobalt 200 virtual machines and outlined additional developments involving its Maia AI accelerators.

The company said Maia 300 is expected to deliver 40% more tokens per watt than leading GPU-based systems. Separately, Microsoft reported that its Maia 200 accelerator is already operating in production to support its largest AI models.

Taken together, the Ignite announcements show Microsoft expanding its own AI model portfolio while continuing to offer third-party systems through Foundry. Its new coding, security and image models emphasize competitive performance at lower operating costs, while the Copilot and Agent 365 updates focus on bringing those capabilities into enterprise workflows.

Ignite 2026 continues through November 20 in San Francisco and online, with Microsoft planning to publish its full collection of announcements following the opening keynote.

This analysis is based on reporting from Testing Catalog.

Image courtesy of Microsoft.

This article was generated with AI assistance and reviewed for accuracy and quality.

Updated Nov 17, 2026

About this article: This article was generated with AI assistance and reviewed by our editorial team to ensure it follows our editorial standards for accuracy and independence. We maintain strict fact-checking protocols and cite all sources.

Word count: 1,042Reading time: 0 minutes

AI News Daily

Breaking Intelligence • Since 2023

Join hundreds of thousands of AI professionals who start their day with our curated newsletter. Get breaking news, expert analysis, and exclusive insights.

Stay Ahead of AI

Get the latest AI breakthroughs, tools, and insights delivered to your inbox every week.

✓ Free forever✓ Unsubscribe anytime✓ No spam guarantee

Go Premium

Unlock unlimited AI tools and an ad-free reading experience designed for AI professionals.

• Ad-free experience• Premium AI tools
Start Free Trial

14-day free trial • Cancel anytime
Plus $9/mo • Pro $90/yr (2 months free)

Follow Our Community

ChatAI

Breaking Intelligence

Your daily briefing on what matters in AI. Trusted by developers, researchers, executives, and AI enthusiasts worldwide.

© 2026 ChatAI. All rights reserved.