OpenAI has introduced an updated ChatGPT Images model that works 4 times faster

OpenAI has begun rolling out a new version of its ChatGPT Images tool, based on GPT-5.2. According to the company, image generation is four times faster, and the model itself executes instructions more accurately, in particular when editing ready-made images – from adding and removing elements to combining and rearranging them.

Particular attention was paid to the quality of text reproduction: the new version works better with inscriptions, including small and dense text, which was a weak point of previous models. A special Images section has appeared in the ChatGPT interface with examples of prompts and ready-made filters for working with graphics.

The update comes against the backdrop of growing competition from the Google ecosystem’s graphics tools, the popularity of which has increased interest in image generation services. OpenAI notes that ChatGPT Images remains an important factor in the growth of the platform’s audience, which has about 800 million active users every week.