Microsoft debuts six AI models as it pushes toward frontier
Add Axios as your preferred source to
see more of our stories on Google.

Photo illustration: Axios Visuals. Photo: Penguin Random House
Microsoft AI CEO Mustafa Suleyman tells Axios the software giant is making significant progress toward building its own frontier general-purpose AI model, even as it introduces six specialized models that it claims rival leading offerings from OpenAI and Anthropic at a fraction of the cost.
Why it matters: Microsoft has long relied on OpenAI and other AI companies for its most advanced models. Now it's demonstrating that it can develop competitive technology of its own, and the new offerings show how aggressively the tech giant plans to compete on price.
Driving the news: Microsoft is announcing six new AI models Tuesday at its Ignite conference in San Francisco. The models span coding, cybersecurity, image generation and voice.
- MAI Code 2: Microsoft says its new coding model delivers 98% of GPT-6.1 Sol's performance at 40% of the price and 96% of Anthropic's Claude Opus 5.5 performance at just 5% of the cost.
- MAI Code 2 Flash: A smaller, faster model that Microsoft says matches GPT-6 Luna at 65% of the price and Claude Haiku 5.5 at 20% of the cost.
- MAI Cyber 2 Flash: A cybersecurity model that Microsoft says beats GPT-6 Astra on a security benchmark at less than 2% of the cost per task.
- Other models: MAI Image 2.7 ranks first in image editing, according to Microsoft, while new transcription and real-time voice models aim to challenge leading rivals.
OpenAI is releasing new models, including Astra and Sol 6.2, which
Between the lines: Microsoft's strategy is to build models specialized for particular tasks, allowing them to achieve comparable performance at significantly lower costs.
"It basically means that you cram more tokens and knowledge and expertise into the same amount of parameters," Suleyman told Axios.
- The trade-off is that a model optimized for everyday coding tasks won't be as capable at advanced math or scientific research.
- Microsoft's performance comparisons rely on selected benchmarks, and some pricing advantages reflect promotional discounts.
The big picture: Even as Microsoft targets the industry's leaders, frontier AI keeps moving forward as well. OpenAI on Tuesday announced GPT‑6.2 Astra and GPT‑6.2 Sol, with the new Astra priced 20 percent less than the prior version.
- Microsoft is also pursuing a frontier general-purpose large language model that can compete across a broader range of tasks.
- Suleyman said Microsoft has achieved state-of-the-art pre-training results and is making strong progress in reinforcement learning, but declined to give a timeline for releasing the model.
- "My team is building true superintelligence," he said.
The intrigue: The new models could reshape Microsoft's relationship with OpenAI and other outside AI providers.
- MAI Code 2 will be available in GitHub Copilot, potentially shifting some coding workloads away from outside models.
- Asked how much Copilot traffic Microsoft expects to move to its own models, Suleyman declined to offer a target but predicted significant adoption.
Zoom in: Microsoft is also offering enterprise customers access to licensed model weights, allowing them to customize its models with proprietary data while requiring deployment within Azure.
- Suleyman said the approach would "feel like open source," while offering greater security and control.
What to watch: How quickly Microsoft can translate its progress in specialized AI into a competitive general-purpose model — and whether its cost advantages are enough to persuade customers to switch from established rivals.
