OpenAI Finally Brings Native Image Generation to ChatGPT

OpenAI Finally Brings Native Image Generation to ChatGPT

After Google released native image generation in Gemini, OpenAI has followed suit and released native image generation to ChatGPT. You can now generate images using the GPT-4o model on ChatGPT. Not only that, you can also upload, edit, and merge multiple images to create a highly consistent image. Thanks to native image generation, you can chat and modify images across multiple generations while maintaining consistency.

OpenAI had demonstrated native image generation with GPT-4o, way back in May 2024. After almost a year, the company has finally added the feature to ChatGPT. Note that instead of using the dedicated Dall -E model to produce images, it leverages the native multimodal capability of GPT-4o to generate and edit images.

  • second image generated using chatgpt native image generation

For example, you can upload an image and ask ChatGPT to create an anime version of the image. You can also use it to generate a series of images, showing visual instructions and demonstrations. Moreover, ChatGPT’s native image generation can be used for producing comic strips, infographics, invitation cards, visual guides, photorealistic images, and much more.

Basically, with native image generation on ChatGPT, you get character consistency across generations, accurate text rendering, and the ability to restyle images. OpenAI also mentions that the native image generation on ChatGPT adheres to instructions very carefully. It can handle up to 10 to 20 different objects in a single prompt.

As for safety, OpenAI says it has given “creative freedom” to users, however, there are robust safeguards against creating sexual deepfakes. The company has also added safety guardrails around nudity and graphic violence. The generated images come with C2PA metadata so people can verify and detect AI-generated images easily.

Coming to availability, ChatGPT’s native image generation is rolling out to all ChatGPT users including free, Plus, Pro, and Team users, starting today. The feature is also available on Sora. Enterprise and Edu users will get the feature pretty soon.

OpenAI Finally Brings Native Image Generation to ChatGPT

Arjun Sha

Passionate about Windows, ChromeOS, Android, security and privacy issues. Have a penchant to solve everyday computing problems.


Source: Beebom