ChatGPT Images 2.0 improves prompt understanding and accuracy

OpenAI 已发布 ChatGPT Images 2.0，这是其图像生成器的全新版本，它改进了对提示的理解能力，并保留了最终输出中的重要细节。

The update comes one year after version 1.0 and introduces “thinking” capabilities that help the system analyze and combine information before creating images. This allows it to produce more accurate and complete visuals based on user input.

ChatGPT Images 2.0 also improves the quality of infographics and diagrams, making them clearer and more accurate, even in different languages. It can also use information from the web to create more context-based visuals.

The tool includes built-in editing features such as background removal, aspect ratio changes, and the ability to generate up to eight images at once in up to 2K resolution. These updates allow users to create and edit images in one place without using separate apps.

ChatGPT Images 2.0 is now available to all users of ChatGPT, while the “Images with thinking” feature is limited to Plus, Pro, Business, and Enterprise plans.