OpenAI makes ChatGPT image generation faster and easier to edit

OpenAI has introduced ChatGPT Images 2.5, a new image generation model for ChatGPT, ChatGPT Work, Codex and the Images API. The company says people now create more than 3 billion images each week across ChatGPT Images and GPT-Image models in the API, so image generation is becoming one of its mainstream product surfaces, not a side feature.

The headline improvement is control. OpenAI says Images 2.5 generates sharper details, more natural lighting and richer textures, preserves subjects from reference photos more reliably, and follows editing instructions better across multiple turns. It also says latency is up to 50% lower than Images 2.0, which matters for iterative creative work where users generate, comment, revise and compare quickly.

ChatGPT is getting workflow features around the model. Sketch lets users draw inside ChatGPT and use that drawing as a visual reference. Templates give starting points for formats such as posters, merch and product photos. Users can also place comments directly on images for targeted edits and share prompts alongside generated images so others can reuse the idea with their own details.

For developers, OpenAI is releasing two API models: GPT-Image-2.5 Flare for most high-volume and product workflows, and GPT-Image-2.5 Sunburst for premium workflows that need tighter control and polished output. OpenAI says the model keeps C2PA metadata and invisible watermarking as part of its safety approach.