Microsoft’s new AI models go beyond just text

Microsoft is doubling down on AI models that aren’t large language models. The company announced Thursday that it will launch three new models: all-new models for voice and text transcription, and the second generation of its in-house image model.

The audio and text transcription models are the first of their kind from Microsoft. The transcription model can translate recordings into text in 25 different languages. It’s built to Video explanation-Meeting transcription and audio agents. The voice model can create audio recordings up to 60 seconds long. The company says its second-generation image model features a faster generation speed and more realistic imaging, resulting in its improvement Its previous model. It is now available in Foundry and Microsoft’s MAI, with future plans to bring MAI-Image-2 to Bing and PowerPoint. Developers can check this Pricing information here.

These new models are a clear sign that Microsoft is looking to expand its offerings across the AI ​​market. Microsoft’s Copilot is one of the most popular chatbots for businesses, especially those already using the Microsoft Office 360 ​​suite and the Azure cloud service. Aside from the now-obsolete original image model, Microsoft has focused primarily on text-based forms, trying to differentiate itself among its many competitors as a secure, enterprise-friendly option. The latest artificial intelligence tools, Co-pilot and Co-pilot healthEvidence of that.

The models are also a reminder that Microsoft, as a legacy technology company, has the money and computing power to exploit these types of “Side quests“Even billion-dollar startups like OpenAI can’t always do this. Last week, OpenAI confirmed that it will Stop the Sora AI video appNoting that it will refocus on core activities. In 2026, the AI ​​industry aims to prove that its tools are useful in the workplace, especially with Claude’s Anthropic Code Jump competition.

It requires generative media, such as models that support AI image and video creation Lots of calculation and energy To run, which can be spent elsewhere. Google, another legacy technology company that has allocated billions of its budget to AI research, indicated this week that it will not abandon generative media but will try to make models more cost and energy efficient, as is the case with its new company. I see a 3.1 lite video model.