Microsoft now implements image creation with GPT-4o in the public version of its AI assistant
After an earlier rollout in the business version of Microsoft 365 Copilot, consumers can now also generate images via GPT-4o in Microsoft Copilot. This makes the utility much more powerful, especially in terms of photorealistic images and accurately rendered text.
Better than Bing Image Creator
GPT-4o can edit and refine existing images via follow-up prompts. These functions elevate Copilot to a higher level than Microsoft Designer and Bing Image Creator, which still use DALL-E 3. Images with correct text, such as posters or labels, are now better achievable.
Thanks to the collaboration with OpenAI, Microsoft continues to benefit from rapid access to new models. Where ChatGPT gets the premiere, Copilot usually follows shortly after. Users also get unlimited access to features such as voice input and in-depth conversations.
A full-fledged AI companion
Microsoft AI chief Mustafa Suleyman previously expressed the ambition to make Copilot a real digital assistant, writes Thurrott. It would remember preferences and be able to perform tasks independently. Meanwhile, Microsoft is also working on its own AI models, which the company says can compete with those of OpenAI and Anthropic.