OpenAI has integrated its most advanced image generator into GPT-4o, making image generation a core capability. This new feature allows users to create images that are both visually appealing and practical, useful for communication and information sharing, not just decoration.
4o image generation has arrived.
— OpenAI (@OpenAI) March 25, 2025
It's beginning to roll out today in ChatGPT and Sora to all Plus, Pro, Team, and Free users. pic.twitter.com/pFXDzKhh2t
The new image generation excels at rendering text accurately, following prompts precisely, and leveraging GPT-4o's knowledge base and chat context. Users can refine images through natural conversation, ensuring consistency across multiple iterations. The model can handle detailed prompts, including rendering between 10-20 different objects, analyze uploaded images, and integrate their details into the generation process. It also uses its knowledge to link text and images.
Available starting today, March 26, 2025, for Plus, Pro, Team, and Free users in ChatGPT, it will soon be available to Enterprise and Edu users, and also in Sora. Developers will gain API access in the coming weeks. All generated images include C2PA metadata to ensure transparency, with internal tools to verify content origin. OpenAI continues to block requests violating content policies and moderates both input text and output images against safety standards.
