ChatGPT now has a free image generator: Here’s how you can use it

OpenAI has just introduced a powerful new feature in ChatGPT: built-in image generation, now available to all users—including those on the free plan. Powered by the multimodal GPT-4o model, this update allows users to generate images directly within the chat interface, eliminating the need to rely on the separate DALL·E 3 model.

What makes GPT-4o image generation different

Unlike its predecessor, GPT-4o uses an autoregressive technique, building images sequentially from top-left to bottom-right. This process results in images with greater visual coherence and far more accurate rendering of complex elements like text and overlapping objects.

Creating images is easier than ever

Users can now ask ChatGPT to generate any kind of image, from infographics to fantasy landscapes, directly in the conversation. Thanks to GPT-4o’s advanced context awareness, the model can remember previous image details, adjust visuals based on user input, and maintain stylistic consistency across iterations.

Powerful tools with real-world applications

From designing marketing materials to creating educational diagrams, GPT-4o’s image generation supports a wide range of use cases. Users can even fine-tune outputs by specifying style, proportions, background transparency, and color palettes with hex codes—bringing professional-level control to anyone, instantly.

Despite its strengths, the system still faces some challenges, like distortion in tall vertical images or difficulty with non-Latin text. All generated images include C2PA metadata and follow strict moderation rules, ensuring responsible and transparent use.