OpenAI Launches ChatGPT Images 2.5 with 50% Faster Generation

OpenAI has launched ChatGPT Images 2.5, a new image model that cuts generation latency by up to 50% while improving detail, editing precision, and reference fidelity, now available to all ChatGPT, ChatGPT Work, and Codex users across desktop, mobile, and web. The release also introduces two new API models, GPT-Image-2.5 Flare and GPT-Image-2.5 Sunburst, targeting developers with different speed and precision trade-offs. The model arrives as users already create more than 3 billion images weekly across ChatGPT Images and GPT-Image models in the API.

September 9, 2026, Inside AI — OpenAI has launched ChatGPT Images 2.5, a new image model that cuts generation latency by up to 50% while improving detail, editing precision, and reference fidelity. The model is now available to all ChatGPT, ChatGPT Work, and Codex users across desktop, mobile, and web, alongside two new API models for developers. The release targets a market where users already create more than 3 billion images weekly across ChatGPT Images and GPT-Image models in the API. Images 2.5 delivers sharper details, more natural lighting, and richer textures. It also preserves subjects from reference photos more accurately and follows multi-turn editing instructions more reliably. New product features include Sketch, which lets users draw directly in ChatGPT as a visual guide, and Templates for popular formats like flyers and product photos. Users can now place comments directly on images for focused edits and share prompts so others can remix ideas with their own photos. Reference Fidelity Anchors Creative Workflows Images 2.5 improves how the model works from reference photos, keeping subjects recognizable across new settings, visual styles, and compositions. Lighting and textures feel more natural, and distinctive features carry through more consistently. For API teams, this fidelity makes reference-led workflows more reliable. Variations stay anchored to the original source, which matters for brand consistency and iterative design. Focused edits are also more precise. The model changes only what the user asks for, preserving other details even with complex subjects and backgrounds. Developers can update a single element, such as a product, background, or piece of copy, while keeping composition and brand treatment intact. Two API Models Split Speed and Precision For developers, OpenAI introduced GPT-Image-2.5 Flare and GPT-Image-2.5 Sunburst. Flare is the default choice for most applications, delivering higher-quality images than GPT-Image-2 at 50% lower latency. It suits creator content, product experiences, visual search, rapid prototyping, and high-volume generation. Sunburst targets premium visual workflows that need tighter control across edits. It is built for production-ready campaign creative or polished product imagery, with longer generation times for extra precision. During longer ChatGPT conversations, Images 2.5 maintains consistency across multiple edits. Earlier changes stay intact, and new edits build on prior work without degrading image quality. This consistency also matters in production workflows where developers make targeted changes without rebuilding entire assets. Images 2.5 better understands complex visual instructions, including transparent backgrounds and accurate real-world content. It reflects visual styles more faithfully, reducing drift as instructions become more specific. That helps generate on-brand creative assets, UI concepts with defined hierarchy, and presentation visuals that fit a set structure. Safety measures build on existing safeguards with checks on prompts and images to prevent harmful outputs. OpenAI continues to use C2PA metadata and invisible watermarking to identify AI-generated images. Pricing details are available on the API platform.