"This has obviously been developed to compete head-on with Google's Nano Banana Pro, in the same way in which GPT-5.2 is a direct response to Gemini 3 Pro. ChatGPT Images "adheres to your intent more reliably", OpenAI's press release says, when you ask it for edits, "down to the small details". It changes only what you ask while keeping everything else intact across inputs, outputs, and subsequent edits."
"This should unlock "results that match your intent", allowing ChatGPT to "act as a creative studio in your pocket, capable of both practical edits and expressive reimaginings". The new model is better at adding, subtracting, combining, blending, and transposing, "so you get the changes you want without losing what makes the image special". Within the new ChatGPT Images feature you'll find preset styles and ideas you can pick without needing to enter a written prompt."
ChatGPT Images is four times faster at generating images and positioned to compete with Google's Nano Banana Pro. The feature adheres to user intent down to small details, changing only requested areas while preserving other elements across inputs, outputs, and subsequent edits. It enables both practical edits and expressive reimaginings by adding, subtracting, combining, blending, and transposing elements without losing what makes the image special. Preset styles and idea selections remove the need for written prompts. Instruction following, text rendering, and the rendering of many small faces are improved. The feature is rolling out globally to all users.
Read at GSMArena.com
Unable to calculate read time
Collection
[
|
...
]