- Enhanced Image Editing – Users can refine AI-generated images through interactive conversations, modifying backgrounds and adding elements for greater customization.
- Improved Text in Visuals – ChatGPT can now generate clearer, structured text in images, making it easier to create diagrams, infographics, menus, and professional materials.
- Gradual Feature Rollout – The new capabilities will be available via the GPT-4o model, with a phased release for free and paid users, as well as software developers.
OpenAI is introducing new features to ChatGPT, expanding its capabilities in image editing and professional visual content creation. The updates, set to be demonstrated in a livestreamed event, will allow users to refine AI-generated images through interactive conversations with the chatbot. This enhancement could make ChatGPT a more valuable tool for both businesses and casual users looking to generate customized graphics.
One of the key improvements is the ability to modify AI-created images with greater precision. Users will be able to request an initial image—such as a snail in an urban setting—then refine details like the background or add elements like accessories. This iterative process allows for more control over the final output, making it easier to create personalized visuals.
Additionally, ChatGPT is improving its text-generation capabilities within images. The AI will now produce clearer, more structured text for elements like diagrams, infographics, and professional materials such as menus and maps. This feature could be particularly beneficial for businesses looking to generate marketing content, presentations, or logos without relying on traditional design software.
The expansion of ChatGPT’s image capabilities aligns with OpenAI’s broader goal of positioning the chatbot as a versatile tool that integrates search functions, voice assistance, and even video generation. These improvements may also help OpenAI stay ahead of competitors, such as Elon Musk’s xAI, which has been developing its own AI-powered image tools.
Despite these advancements, ChatGPT still faces challenges, particularly in handling text accuracy within images. Errors such as generating fake country names or struggling with small-sized text and non-Latin alphabets remain potential issues. OpenAI has acknowledged these limitations, emphasizing that user input quality plays a significant role in the final output. The new features will be available through the GPT-4o model, with a gradual rollout planned for both free and paid users, as well as software developers accessing OpenAI’s API.
