OpenAI continues its 'code red' roll-out with faster image creation
OpenAI is stepping up its fight with Google in image generation with a faster and more capable version of its picture making tool for ChatGPT.
The new system, called GPT Image 1.5, is being rolled out to all ChatGPT users and to developers through the API. OpenAI says it can turn prompts into images up to four times faster than before and is better at following detailed instructions and making precise edits.
The launch comes soon after a leaked internal memo from chief executive Sam Altman reportedly told staff the company was on a “code red” footing as it tries to reclaim ground from Google. The search group has been winning attention with its latest Gemini 3 model and its Nano Banana Pro image generator, which have been scoring highly on independent benchmarks. Even after the release of OpenAI’s GPT-5.2 language model last week, Google has kept a lead in some rankings.
OpenAI had originally planned to introduce a new image model early next year but has brought the timetable forward. GPT Image 1.5 follows on from GPT Image 1, which arrived in April, and reflects how generative image and video tools are shifting from novelty status towards everyday production use in areas such as marketing, design and entertainment.
One of the main upgrades is in post-production. Many image tools struggle when a user wants to tweak a single element. Ask for a small change, such as a cooler light or a different facial expression, and the model often redraws the whole scene. OpenAI says GPT Image 1.5 is better at making targeted adjustments while keeping the rest of the image stable. That includes handling things such as faces, lighting, composition and colour so that edits stay consistent from one version to the next.
There is also a change in how people access and work with images inside ChatGPT. Pictures now sit behind a dedicated button in the sidebar that is meant to feel more like a simple creative studio. New viewing and editing screens are designed to make it easier to refine an image, try out variations and browse popular prompts and preset filters for ideas.
Alongside the new generator, OpenAI is trying to make ChatGPT itself more visual. Search style queries are expected to show more pictures and diagrams, with clear sourcing, for tasks where a graphic answer is more useful than plain text. Examples include unit conversions, quick look ups and live information such as sports scores.
The broader aim is to narrow the gap between an initial idea in the user’s head and the finished result on screen. By speeding up image creation, improving editing and weaving visuals more deeply into ChatGPT, OpenAI is responding to pressure from Google and other rivals. For investors watching the AI race, it is another sign that image tools are moving to the heart of the competition between the big model providers.
