You can now use Microsoft's first in-house AI image generator
What's the story
Microsoft has launched its first in-house artificial intelligence (AI) image generator, MAI-Image-1. The model is now integrated into two of the company's products: Bing Image Creator and Copilot Audio Expressions. Mustafa Suleyman, the head of Microsoft AI, announced this development on X. He also revealed that the text-to-image model will be available soon in the European Union (EU).
Model capabilities
What is MAI-Image-1 good at?
Suleyman highlighted the strengths of MAI-Image-1, saying it is especially good at creating images of food and nature scenes. The model also does artsy lighting and photorealistic detail really well. As per the company, MAI-Image-1 is faster and better than many other image models. "Its combination of speed and quality means users can get their ideas on screen faster, iterate through them quickly," Microsoft said in its blog post.
Feature integration
MAI-Image-1 can also create AI art
MAI-Image-1 can also be used to create art for AI-generated audio stories in the "story mode" of Copilot's text-to-speech platform, Copilot Audio Expressions. The move comes after Microsoft's earlier announcement of its first in-house AI models - speech model MAI-Voice-1 and text-based model MAI-1-preview. These developments hint at a possible shift away from OpenAI's models.
Model availability
MAI-Image-1 joins DALL-E 3 on Bing's image creator
MAI-Image-1 is one of the three AI models listed on Bing's image creator website and app. The other two are DALL-E 3 and GPT-4o from OpenAI. This further shows Microsoft's commitment to developing its own in-house AI capabilities while still leveraging existing technologies from other industry leaders like OpenAI.