OpenAI has announced the release of ChatGPT Images 2.5, a state-of-the-art image generation model designed to enhance detail, texture, and lighting accuracy. A significant technical improvement is the reduction of image generation latency by up to 50% compared to the previous 2.0 version. The new model demonstrates superior performance in preserving subject identity when working from reference photos and exhibits higher reliability in multi-turn editing workflows, maintaining image consistency and quality across successive modifications. For users within ChatGPT, the update introduces "Sketch," allowing for direct drawing as a visual reference, alongside new templates for specific formats like flyers and product photos, and the ability to place targeted comments on images for precise editing. For developers, the Images API has been expanded to include two new models: GPT-Image-2.5 Flare, which offers high-quality output with the aforementioned 50% latency reduction, and GPT-Image-2.5 Sunburst, which is optimized for premium, high-precision creative workflows requiring intensive control. To ensure safety and provenance, the system continues to utilize C2PA metadata and invisible watermarking to identify AI-generated content. The rollout is available across all ChatGPT tiers and via the API.