Microsoft launches first in-house AI image model challenging Google, OpenAI
What's the story
Microsoft has officially unveiled MAI-Image-1, its first-ever in-house text-to-image generator. The tech giant describes this innovative tool as "the next step on our journey." The development of MAI-Image-1 was heavily influenced by feedback from creative professionals who wanted to avoid "repetitive or generically-stylized outputs."
Advanced features
MAI-Image-1's capabilities and performance
Microsoft claims that MAI-Image-1 excels at creating photorealistic images of lightning, landscapes, and more. It is also faster than "larger, slower models" in processing requests and generating images. The model has already made its mark by entering the top 10 on LMArena, an AI benchmark site where human users compare outputs from different systems and vote on the best one.
Product expansion
Microsoft's AI advancements and model investments
MAI-Image-1 is the latest addition to Microsoft's growing suite of AI products, which already includes voice generator MAI-Voice-1 and chatbot MAI-1-preview. The company was an early investor in OpenAI, but their relationship has become increasingly complicated over time. Recently, Microsoft has also started using Anthropic's AI models for some features in Microsoft 365 and is making "significant investments" to train its own AI models like MAI-Image-1.