Microsoft unveils two new AI models, signalling direct competition with OpenAI

Microsoft unveils two new AI models, signalling direct competition with OpenAI

In a significant move to bolster its presence in the artificial intelligence sector, Microsoft unveiled two groundbreaking AI models on Thursday: MAI-Voice-1 and MAI-1 Preview. Historically, the tech giant has leaned heavily on OpenAI's GPT models to enhance its Copilot features. However, with these new releases, Microsoft is positioning itself as a direct competitor to OpenAI and other key players in the AI landscape. The MAI-Voice-1 model is designed for natural speech generation and boasts the impressive capability of producing a full minute of audio in less than a second, all while utilizing just a single GPU. This efficiency makes it one of the most advanced speech models available, now integrated into the Copilot Daily and Podcast features, along with its introduction in Copilot Labs. On the other hand, MAI-1 Preview serves as a foundational AI model, which is currently undergoing public testing on LMArena, a platform dedicated to benchmarking large language models. Although it currently holds the 13th spot on the LMArena leaderboard, trailing behind notable models such as GPT-5 and Gemini 2.5 Pro, Microsoft's approach to model training sets it apart. The MAI-1 Preview has been trained using 15,000 NVIDIA H100 GPUs, a fraction of the more than 100,000 GPUs used by competitors like Grok. In an interview with Semafor, Mustafa Suleyman, Microsoft’s AI Chief, emphasized the company's commitment to leveraging insights from the open-source community to maximize the capabilities of its models while minimizing resource usage. He stated, "The art and craft of training models now lies in selecting the ideal data and optimizing resources, ensuring that every computational effort contributes meaningfully to the model's learning process." Looking ahead, Suleyman revealed that Microsoft AI is already developing the next iteration of its AI models, with plans to utilize some of the largest data centers globally, equipped with NVIDIA's latest GB200s chips. He expressed optimism about the future, stating, "We have ambitious plans moving forward. Our goal is not only to achieve further advancements but also to create a suite of specialized models tailored to various user intents and applications, unlocking tremendous value in the process. Expect more exciting developments from our team soon."

Sources : Mint

Published On : Aug 29, 2025, 05:10

AI
Anthropic Takes Legal Stand Against Government's Supply Chain Risk Label

Dario Amodei, CEO of Anthropic, has announced that his company will be legally contesting the U.S. government's recent c...

CNBC | Mar 06, 2026, 01:05
Anthropic Takes Legal Stand Against Government's Supply Chain Risk Label
AI
Anthropic Downplays Pentagon's Supply Chain Risk Designation Amid Ongoing Negotiations

On Thursday, Anthropic addressed concerns regarding the Pentagon's classification of the AI firm as a 'supply chain risk...

CNN | Mar 06, 2026, 02:50
Anthropic Downplays Pentagon's Supply Chain Risk Designation Amid Ongoing Negotiations
AI
India's GCCs Transforming into AI-Driven Powerhouses Amid Evolving Policies

India's global capability centres (GCCs) are undergoing a significant transformation, shifting from traditional cost cen...

Business Today | Mar 06, 2026, 09:50
India's GCCs Transforming into AI-Driven Powerhouses Amid Evolving Policies
Cybersecurity
Karnataka Takes Bold Step: Social Media Ban for Under-16s Announced

In a significant move to safeguard children, Karnataka Chief Minister Siddaramaiah announced on March 6 a statewide proh...

Business Today | Mar 06, 2026, 08:05
Karnataka Takes Bold Step: Social Media Ban for Under-16s Announced
Cybersecurity
Meta's AI Smart Glasses Under Fire: Lawsuit and Regulatory Scrutiny Over Privacy Violations

Meta's AI-enhanced Ray-Ban smart glasses are currently at the center of a legal storm due to serious privacy concerns. R...

Business Today | Mar 06, 2026, 06:05
Meta's AI Smart Glasses Under Fire: Lawsuit and Regulatory Scrutiny Over Privacy Violations
View All News