- Today
- Holidays
- Birthdays
- Reminders
- Cities
- Atlanta
- Austin
- Baltimore
- Berwyn
- Beverly Hills
- Birmingham
- Boston
- Brooklyn
- Buffalo
- Charlotte
- Chicago
- Cincinnati
- Cleveland
- Columbus
- Dallas
- Denver
- Detroit
- Fort Worth
- Houston
- Indianapolis
- Knoxville
- Las Vegas
- Los Angeles
- Louisville
- Madison
- Memphis
- Miami
- Milwaukee
- Minneapolis
- Nashville
- New Orleans
- New York
- Omaha
- Orlando
- Philadelphia
- Phoenix
- Pittsburgh
- Portland
- Raleigh
- Richmond
- Rutherford
- Sacramento
- Salt Lake City
- San Antonio
- San Diego
- San Francisco
- San Jose
- Seattle
- Tampa
- Tucson
- Washington
Redmond Today
By the People, for the People
Microsoft Unveils Three New AI Models for Voice and Images
The tech giant's internal AI research group launches foundational models to compete with OpenAI, Google, and others.
Apr. 5, 2026 at 6:05pm
Got story updates? Submit your updates here. ›
Microsoft has introduced three new large-scale AI models focused on speech-to-text transcription, audio generation, and image creation. These foundational models, developed by the company's internal MAI research group, position Microsoft to reduce its reliance on OpenAI and compete more directly with other major tech players in the rapidly evolving AI landscape.
Why it matters
By building its own AI models, Microsoft gains greater control over its technology strategy and the ability to tailor these systems to its own products and infrastructure. This move reduces the company's dependence on external partners like OpenAI, whose technology currently powers Microsoft's Copilot AI assistant.
The details
The three new MAI models cover key media types: speech-to-text transcription, audio generation from prompts, and image creation based on text descriptions. These foundational models are designed to be integrated into Microsoft's apps, services, and devices, allowing the company to optimize performance and costs compared to licensing external AI technology.
- Microsoft's internal MAI research group launched about six months ago.
- The three new models represent MAI's first significant public release.
The players
Microsoft
A multinational technology company headquartered in Redmond, Washington, known for products like Windows, Office, and Azure cloud services.
MAI
Microsoft's internal AI research group that developed the three new foundational AI models.
OpenAI
An artificial intelligence research company that has provided technology powering Microsoft's Copilot AI assistant.
What they’re saying
“Microsoft building their own models is huge. They've been paying OpenAI basically to compete against themselves in some markets. Makes zero sense long-term.”
— Reddit user
“Okay but can we see actual benchmarks before getting excited? Everyone says their model is competitive until you actually test it.”
— YouTube commenter
What’s next
Independent testing of MAI's models against competitors like OpenAI Whisper, Google Imagen, and others will be crucial to evaluating their performance. Additionally, any updates to Microsoft's Azure cloud platform or the company's Copilot AI assistant that integrate the new MAI models will signal how quickly the technology is being deployed.
The takeaway
Microsoft's development of its own foundational AI models represents a strategic shift to reduce reliance on external partners and gain more control over its technology ecosystem. This move positions the company to better compete with rivals like Google and OpenAI in the rapidly evolving AI landscape, potentially leading to improved capabilities across Microsoft's products and services.
