ChatGPT image generation is now faster and better at following tweaks (www.engadget.com)

OpenAI has announced the rollout of an upgraded image generation model within ChatGPT, following the recent release of GPT-5.2. This new version is reported to be four times faster than its predecessor, addressing delays that users often experienced, especially during peak times. The model also enhances instruction adherence, allowing for more nuanced edits such as adding, subtracting, and blending elements within images. A key improvement is its text rendering capabilities, now better equipped to handle denser and smaller text, a notorious challenge for image models. This update is significant for the AI and machine learning community as it strengthens OpenAI's competitive edge against rivals like Google’s Gemini, which has seen rapid user growth. The introduction of a dedicated Images section in the ChatGPT sidebar, featuring preset filters and prompts, further enhances usability and creativity. OpenAI emphasizes that this is just the beginning of advancements in image generation, with expectations for more detailed outputs and refined edits in future updates. As the rollout begins today, existing users can still access the previous model, ensuring a seamless transition for all.