SecureITWorld (1)
Sign Up

We'll call you!

One of our agents will call you. Please enter your number below

JOIN US



Subscribe to our newsletter and receive notifications for FREE !





    By completing and submitting this form, you understand and agree to SecureITWorld processing your acquired contact information as described in our Privacy policy. You can also update your email preference or unsubscribe at any time.

    SecureITWorld (1)
    Sign Up

    JOIN US



    Subscribe to our newsletter and receive notifications for FREE !





      By completing and submitting this form, you understand and agree to SecureITWorld processing your acquired contact information as described in our Privacy policy. You can also update your email preference or unsubscribe at any time.

      Microsoft AI Unveils Seven ‘World-Class Models’ At Comparatively Lower CostsĀ 

      MAI's Seven Models

      Microsoft's latest AI strategy places affordability, specialized capabilities, and operational efficiency at the center of model development. For this purpose, the company's MAI has introduced seven AI models designed for coding, cybersecurity, image generation, voice, and business decision-making. All of them come at a lower cost than other available models at present.

      The launch of MAI's seven models signals Microsoft's growing efforts to make sophisticated AI capabilities available without costly infrastructure requirements. The step also strengthens the company's position in an increasingly competitive AI development environment.

      MAI's Seven Models Target Different Workloads

      MAI focuses on addressing diverse workloads within an organization with the launch of seven models. The models allow users to run agents 24/7 doing real engineering work. Microsoft released the models at its annual conference, Microsoft Ignite 2026, held at the Moscone Center in San Francisco from November 17-20, 2026. MAI's Seven Models are-

      According to Microsoft, MAI Code 2 achieves 98% of GPT-6.1 Sol's coding performance at approximately 40% of its cost.Ā  The model delivers 96% performance for approximately 5% of the price compared to Claude Opus 5.5.

      Similarly, Code 2 Flash targets frequent coding activities, including routine fixes, targeted edits, explanations, and faster development iterations. Microsoft positions the model as a more economical option for high-volume programming workloads.

      Specialized Models for Cybersecurity and Image Generation

      MAI Cyber 2 Flash focuses on vulnerability discovery, threat investigation, and recommending appropriate fixes for cybersecurity applications. Microsoft reports a 92.3% CyberGym-E2E score, compared with GPT-6 Astra's 91.5%. The company also claims that the model completes these benchmark tasks at less than 2% of the competing model's cost.

      On the other hand, MAI Image 2.7 and its Flash variant expand Microsoft's offering into image generation and editing. The flagship model ranks first in Artificial Analysis image editing evaluations, according to Microsoft's announcement. The model's priced approximately at one-fifth of the cost of GPT Image 2.6 Sunburst.

      The Lineup Includes Models for Voice, Transcription, and Decision-Making

      Microsoft's MAI Realtime 1 introduces its first speech-to-speech model for natural enterprise conversations for voice applications. The model handles interruptions, background noise, and requests requiring connections with external business systems. The model reportedly outperforms GPT-Realtime 2.1, with pricing starting at $0.01 per audio minute.

      The transcription offering, MAI Transcribe 2, already supports Microsoft's Dragon Copilot healthcare platform. The model is capable of processing approximately 10 million patient encounters monthly while reducing multilingual transcription error rates by half.

      Microsoft-Decision-1.1 completes the lineup with specialized capabilities for routing, classification, prioritization, and workflow control. The model operates 50 times faster and costs 79 times less than GPT-6.1 Sol. It is now generally available through Microsoft Foundry, OpenRouter, and Vercel.

      Microsoft's Initiative to Expand Enterprise Control Over AI Models

      Microsoft is introducing MAI Licensed Weights for enterprise customers looking for greater control, rather than focusing only on performance and pricing. Organizations can therefore customize eligible models using proprietary data and deploy them securely within their Azure environments. When it comes to model control, Microsoft claims that customers retain control over access, data, and model adaptations.

      The announcement of seven models at once reflects Microsoft's strategy of matching specialized intelligence with specific business requirements. The company prioritizes practical capabilities and lower operating expenses alongside pursuing maximum performance across diverse tasks within an organization.

      SecureITWorld is your one-stop destination for all your tech and cybersecurity needs. Visit our site and start learning to make informed decisions.


      Also Read:

      Zero Trust for AI: Microsoft’s Approach to Secure the Full Lifecycle of AI

      Microsoft Scout Explained: The Always-On Autonomous Agent for Your Workflows





        By completing and submitting this form, you understand and agree to SecureITWorld processing your acquired contact information as described in our Privacy policy. You can also update your email preference or unsubscribe at any time.

        Popular Picks


        Recent Blogs

        Recent Articles

        SecureITWorld (1)

        Contact Us

        For General Inquiries and Information:

        For Advertising and Partnerships: 


        Copyright Ā© 2026 SecureITWorld . All rights reserved.

        Scroll to Top