OpenAI brings ChatGPT Images 2.5 with faster generation and improved image editing

OpenAI has released ChatGPT Images 2.5, a new image-generation model designed to deliver sharper details, faster creation and more precise editing. The update improves natural lighting, richer textures and reference-photo fidelity, while making multi-turn edits more consistent.

Must Read: AI security: How India is using technology to protect its expanding mobility infrastructure

OpenAI says Images 2.5 can follow complex visual instructions more reliably, including layouts, transparent backgrounds and artistic styles. Alongside the model, ChatGPT gets new tools such as Sketch, templates, image comments and prompt sharing.

The changes mean more control when creating or refining visuals, while developers get two new API models aimed at different creative workflows and tasks.

ChatGPT Images 2.5: Upgrades and new creative tools

Images 2.5 preserves distinctive features from reference photos across settings, styles and compositions. It can make focused edits without unnecessarily changing surrounding details. Across longer conversations, earlier changes are more likely to remain consistent.

Must Read: OpenAI says ChatGPT can now mimic your writing style from emails and messages; Here’s how

OpenAI says generation latency is up to 50 per cent lower than Images 2.0. The model handles real-world information more accurately and supports complex layouts, including transparent backgrounds.

In ChatGPT, Sketch lets you draw an idea as a visual guide - type “@Sketch” to try it. Templates help users start with formats such as posters and merchandise. Comments can be placed on images for focused edits, while shared prompts let others recreate an idea with their photos.

ChatGPT Images 2.5 API models and availability

OpenAI is introducing GPT-Image-2.5 Flare and GPT-Image-2.5 Sunburst in its API. Flare brings quality, editing and speed improvements, with 50 per cent lower latency than GPT-Image-2, for creator and social content, product experiences, visual search and high-volume generation. Sunburst targets workflows requiring tighter editing control, including campaign creative and polished product imagery.

People create over 3 billion images weekly across ChatGPT Images and GPT-Image models in the API. Images 2.5 is rolling out to ChatGPT, ChatGPT Work and Codex users across desktop, mobile and web. Safeguards include prompt and image checks, C2PA metadata, and invisible watermarking.