Do you use ChatGPT to generate images? Then you’ll be happy that OpenAI has now added a new image generator.
Creating images with ChatGPT now works better
OpenAI has a new image generator called GPT Image 1.5 added to ChatGPT. This allows ChatGPT to generate images up to four times faster and perform precise edits to an image while retaining details. OpenAI says ChatGPT can now adjust only the details you want to update, leaving lighting, composition, and people’s appearance unchanged.
The new model excels at adding, removing, combining, mixing and transposing real photos for more realistic image creations. GPT Image 1.5 is also better at adding text to images and changing layouts. OpenAI says the new model follows instructions more reliably than the previous generation. So, for example, you can ask for the people in the photo to wear different clothes. Or to add something in the background. ChatGPT then only changes what you requested and leaves everything else unchanged.
More changes
With the update, the ability to generate certain specific art styles such as anime has deteriorated. ChatGPT’s new model is also less good at changing images with many people. According to OpenAI, using preset filters can help with this. ChatGPT Images contains dozens of preset styles and prompts that you can try out on images. Moreover, the previous version of the image generator is also still available for use. You can use the new version in the mobile app and also on the ChatGPT website.
The updated Images model is currently rolling out to all ChatGPT users. Thanks to this new image generator, ChatGPT can compete much better with the Nano Bananaimage generator from Google.
Don’t want to miss out on all the developments in image generation with ChatGPT? Then sign up for our newsletter and we will automatically keep you informed from now on.
Download the iPhoned app
