OpenAI has released ChatGPT Images 2.5, a new image-generation model the company says brings sharper detail, faster output and more precise editing across its consumer and developer products.
Faster Generation, Better Fidelity
OpenAI said more than three billion images are created weekly across ChatGPT Images and the GPT-Image models available through its API.
With Images 2.5, the company said generation latency has been cut by up to 50% compared with the previous Images 2.0 model, while output has more natural lighting and richer textures.
The model is also better at preserving subjects from reference photos, according to OpenAI, retaining recognisable features, lighting and textures when a person, place or object is placed into a new setting, style or composition.
Sharper Editing Controls
The company said Images 2.5 follows editing instructions more reliably across multiple turns of a conversation, allowing users to make several rounds of changes without the image losing quality or drifting from earlier edits.
It also handles more complex visual instructions, including layered compositions and transparent backgrounds, and is better at matching a requested artistic style, which OpenAI said makes it more dependable for creative briefs that specify a particular tone, hierarchy or brand treatment.
ALSO READ: OpenAI Restores 5-Hour Codex, Work Limits For ChatGPT Plus Users
New Creative Tools
Alongside the model, OpenAI introduced product features aimed at giving users more control over their output.
A tool called Sketch lets people draw directly within ChatGPT to guide the composition of a final image, while new templates are designed to simplify popular formats such as flyers and product photos.
ALSO READ: AI Race Tightens: Gemini, Claude Eat Into ChatGPT's Global User Share. Here's What Report Says
Users can also place comments directly on images for more targeted edits and share prompts alongside their own photos so others can build on them.
OpenAI said Images 2.5 is available to all ChatGPT, ChatGPT Work and Codex users across desktop, mobile and web.
For developers, the company is introducing two new API models: GPT-Image-2.5 Flare, which carries the same improvements in quality, editing and speed, and GPT-Image-2.5 Sunburst, which offers an additional level of precision for detailed creative work at longer generation times.
Essential Business Intelligence, Sharp Market Insights, Practical Personal Finance Advice, Daily Fuel, Gold and Silver Prices and Latest Stories — On NDTV Profit.
