Microsoft AI (MAI) has introduced its proprietary AI models, MAI-Voice-1 and MAI-1 Preview, stepping into a landscape where it has mainly depended on OpenAI’s GPT models.
With this launch, Microsoft directly enters the competitive AI market alongside OpenAI.
Access deeper industry intelligence
Experience unmatched clarity with a single platform that combines unique data, AI, and human expertise.
MAI-Voice-1, available in Copilot Daily and Podcasts, is designed to produce high-fidelity and expressive audio suitable for both single and multi-speaker scenarios. This model allows users to interact with AI through voice, which Microsoft sees as a crucial interface for future AI companions.
Users can explore this technology further through Copilot Labs, where they can experiment with expressive speech and storytelling demos.
MAI-Voice-1 is said to stand out for its efficiency, capable of generating a full minute of audio in under a second on a single GPU. Microsoft claimed that this makes the model one of the fastest speech generation systems available.
The other model, MAI-1 Preview, represents Microsoft’s initial attempt at developing an in-house foundation model trained end-to-end. It is currently undergoing public testing on LMArena, a platform dedicated to community evaluations of AI models.
US Tariffs are shifting - will you react or anticipate?
Don’t let policy changes catch you off guard. Stay proactive with real-time data and expert analysis.
By GlobalDataThis model is a mixture-of-experts designed to enhance capabilities in instruction-following and responding to user queries effectively. It was trained on approximately 15,000 NVIDIA H100 GPUs, enabling it to handle complex tasks and deliver robust performance.
In the coming weeks, Microsoft plans to integrate MAI-1 Preview into selected text use cases within Copilot to gather user feedback and make iterative improvements.
The company said that it is committed to leveraging insights from its internal teams, partners, and open-source innovations to enhance product performance across numerous interactions daily.
Microsoft AI, in a blog, stated: “We have big ambitions for where we go next. Not only will we pursue further advances here, but we believe that orchestrating a range of specialised models serving different user intents and use cases will unlock immense value.
“There will be a lot more to come from this team on both fronts in the near future.”
