OpenAI has launched GPT Image 1.5, a new AI model focused on image generation and editing. This model arrives shortly after Google’s launch of Nano Banana Pro, which improves text rendering and instruction-following capabilities. GPT Image 1.5 also enhances these areas, along with optimizing processing efficiency. Developers can access it through a new API that costs 20% less than the previous version. For nontechnical users, image generation is available via ChatGPT, which now operates up to four times faster. One significant improvement in GPT Image 1.5 is its ability to retain important image details during edits. This is particularly beneficial for commercial users who need precise adjustments, like resizing product images without altering logos. The model is also capable of executing complex edits, such as combining elements from multiple images and changing their artistic styles. OpenAI has focused on text rendering as well, making it more effective for creating infographics and other text-heavy visuals. Despite its advancements, GPT Image 1.5 has limitations, such as limited support for certain drawing styles and occasional errors in generating scientifically accurate images. However, OpenAI claims that it performs better in these areas compared to its predecessor. The model is integrated into ChatGPT, featuring pre-configured filters and prompts that will be updated regularly. This launch coincides with the recent upgrade of ChatGPT to the GPT-5.2 foundation model, which has shown improved performance in scientific tasks.