OpenAI launches ChatGPT Images 2.0 with improved text rendering: Availability, price and more - The Times of India

OpenAI has released ChatGPT Images 2.0, a state-of-the-art upgrade to its image generation technology, designed to tackle complex visual tasks and produce highly precise and immediately usable images. This new model marks a considerable leap from its predecessor, particularly in its ability to accurately render small text, icons, and user interface elements, which has historically posed a challenge for AI image generators. Key advancements include improved instruction following, accurate object placement, and the capability to generate visuals across various aspect ratios. The model's enhanced "sense of composition and visual taste" contributes to images that appear less artificial and more intentionally designed. Furthermore, ChatGPT Images 2.0 offers superior multilingual support, accurately generating text within images in languages such as Japanese, Korean, Chinese, Hindi, and Bengali, thereby facilitating the creation of diverse visual content like posters and diagrams with better readability and structure. OpenAI states that the model leverages expanded visual and world knowledge to generate smarter images with less prompting.