OpenAI integrates AI image generation to ChatGPT and Sora

OpenAI's long-awaited improved AI image-generation capabilities is now available. Rather than using a separate picture generation model like Dall-E, the new enhanced image generator is now included into GPT-4o.

There are various AI image-generation models on the market that can produce bizarre, gorgeous images. They all struggle to generate graphics that include text, logos, and other basic elements found in everyday life.

OpenAI believes that the new GPT-4o picture generation may address these issues by accurately rendering text and precisely following directions using its knowledge base and chat context. This new approach also allows users to edit uploaded photographs or use them as visual inspiration to generate new ones.

The updated GPT-4o image-generation model is now available to all ChatGPT Plus, Pro, Team, and Free users. Since this new model will be the default image generator in ChatGPT, users will be able to enjoy it without having to make any further selections before their prompt. Users can additionally customise their photographs by specifying the aspect ratio, specific colours using hex codes, or a transparent background. OpenAI will also make this new model available to ChatGPT Enterprise and Edu users in the coming weeks.

This new model is also available in Sora for picture creation and through the specific DALL·E GPT. Image generation with the GPT-4o API will become available to developers in the next weeks.

All photographs generated by this new model will have C2PA metadata, and OpenAI's internal tool will be able to determine whether an image was created using this model.