At Ignite on November 17, Microsoft introduced a family of seven new MAI AI models that it says come close to leading rivals at a fraction of their price.
The headline model is MAI Code 2, aimed at everyday enterprise coding. Microsoft says it delivers 98% of GPT-6.1 Sol’s coding quality at about 40% of the price, and 96% of Claude Opus 5.5’s performance at about 5% of the price. It is available in GitHub Copilot now, while the lighter MAI Code 2 Flash targets routine edits and explanations in Copilot and in a limited preview on Microsoft Foundry.
MAI Image 2.7 took first place in Artificial Analysis’s image-editing ranking, according to Microsoft, and its Flash variant generates images for about $10 per 1,000. MAI Realtime 1, the company’s first live speech-to-speech model, starts at $0.01 per audio minute. MAI Cyber 2 Flash, built for security work, is said to beat GPT-6 Astra’s quality on the CyberGymE2E benchmark at under 2% of the cost per task.
Microsoft also launched MAI Licensed Weights, letting companies customise MAI models with their own data and run them inside their own Azure environment, and said its Maia 200 AI accelerator is now in production with 1.7 times more tokens per watt.
Why it matters
Microsoft is building its own cheaper alternative to its partners’ models, adding pressure on AI pricing across the industry. For governments and companies already on Azure, including many in the Gulf and wider Arab region, running a customised model inside their own environment could ease data-sovereignty concerns. All performance and price figures come from Microsoft and still await independent testing.
Source: Microsoft AI
By the TechMojaz team

اترك تعليقاً