Microsoft is leaning hard into price as its weapon in the AI model race. At Ignite 2026 on Tuesday, Microsoft AI released a family of seven new MAI models covering coding, cybersecurity, images, voice, transcription and model routing, each pitched as offering strong performance at a much lower cost than rival frontier models.
“As token spend skyrockets, we know that customers also want models at a price point that lets you run agents 24-7 doing real engineering work,” Microsoft AI wrote in its announcement.
The coding models
The centerpiece is MAI Code 2, built for everyday enterprise coding. Averaged across four benchmarks, Microsoft says it delivers 98% of the performance of OpenAI’s GPT-6.1 Sol at about 40% of the price, and 96% of Claude Opus 5.5’s performance at just 5% of the price. Microsoft is upfront that the models are “explicitly not yet tuned” for harder problems like fundamental science research or very long-horizon agentic work.
MAI Code 2 is available in GitHub Copilot today. MAI Code 2 Flash, aimed at high-volume routine tasks, is in GitHub Copilot and in limited preview in Microsoft Foundry.
Security, images and voice
MAI Cyber 2 Flash, Microsoft says, matches GPT-6 Astra’s quality at less than 2% of the cost per task on the CyberGymE2E benchmark, and is priced 86% lower on output tokens than MAI Cyber 1.1 Flash.
The lineup also includes MAI Image 2.7, which Microsoft says ranks first on Artificial Analysis for image editing; MAI Realtime 1, a speech-to-speech model starting at $0.01 per audio minute; Dragon Copilot Transcribe for clinical transcription; and Decision 1.1, a routing model that sends requests to bigger, pricier models only when needed. Microsoft also says its second-generation Maia 200 AI accelerator is now in production.
Your weights, your cloud
The most interesting enterprise piece may be MAI Licensed Weights. Customers can take a frontier MAI model, customize it with proprietary data and run it in their own Azure environment, controlling who can access it. Microsoft cited AT&T, which built a custom MAI model for 5G network operation analytics that it says outperformed GPT-6.1 Sol at 50x lower cost.
Why it matters
As Axios frames it, Microsoft is now taking on OpenAI and Anthropic with its own models, and it is competing on cost. For enterprises running agents around the clock, a model that is “good enough” at a fraction of the price can change the math on AI projects.
For software vendors, the message is that AI cost per task is becoming a buying criterion. If your product runs on large models, be ready to explain your model choices, routing and pricing. Buyers comparing tools will increasingly ask what they’re paying for each outcome, not just which model sits underneath.
Source: Microsoft AI · Axios. Written by the xRapid Media LLP team.
About xRapid Media
xRapid Media LLP is a High Tech B2B Marketing company for technology businesses: software and SaaS companies, IT services and managed service providers, and hardware and IT product vendors. Our team combines positioning and content, account-based marketing, paid ads funnels, verified B2B data, intent signals, email and LinkedIn outreach and appointment setting to help you reach the right buyers and turn their attention into qualified leads and meetings.
Want help putting ideas like these into practice? Email [email protected] or book a time with our team.