OpenAI says the new ChatGPT Images can make more precise edits and have higher instruction following.
Developers can access the new image model dubbed GPT Image 1.5 via API
OpenAI has launched an upgraded version of ChatGPT Images, a major overhaul of the company's visual generation feature. The new tool is powered by a flagship image generation model called GPT Image 1.5, bringing major improvements in instruction-following, editing precision and generation speed. The San Francisco-based artificial intelligence (AI) giant said image creation and iterations are now up to four times faster than before. Interestingly, these improvements make ChatGPT Images comparable to Google's Nano Banana Pro, which was also released recently.
In a post, OpenAI announced and detailed the upgraded ChatGPT Images feature. With this iteration, the company has made the tool more responsive to instructions. Users can describe what they want, and the tool will produce results that match those instructions more closely, whether they are creating visuals from scratch or using other photos. The precision also comes in handy in editing tasks where unwanted changes can now be mitigated.
One of the biggest changes with the new ChatGPT Images is speed. OpenAI says the updated model can generate and edit images up to four times faster than before. This can matter a lot when people are iterating on ideas, such as testing variations on a design or refining a photo edit, because less waiting time means a more efficient experience.
Under the hood, these capabilities are powered by a new flagship image generation model, which replaces previous bases for image generation. OpenAI says it now follows user intent more reliably, and when making edits, it only changes the requested area and can now keep more abstract elements, such as lighting, composition, and people's appearance, consistent across multiple edits.
OpenAI has also added a dedicated Images interface within ChatGPT. This separates image generation from general chat interactions and provides users with a clearer space to explore visual ideas. The space can be accessed from the sidebar in the app and website. Users can find a large number of preset filters that they can tap, upload an image, and make an edit in that style without having to type long and highly detailed prompts. The company says the space also acts as an inspiration zone.
ChatGPT Images is now rolling out to all users, including those on the free tier, globally. Additionally, developers can also select the AI model by selecting GPT Image 1.5 in the application programming interface (API). Notably, the older version of the image tool will be available as a custom GPT.
Get your daily dose of tech news, reviews, and insights, in under 80 characters on Gadgets 360 Turbo. Connect with fellow tech lovers on our Forum. Follow us on X, Facebook, WhatsApp, Threads and Google News for instant updates. Catch all the action on our YouTube channel.
Baldur's Gate 3 Developer Larian Studios Says It Uses Generative AI, CEO Responds to Backlash
Honor Win, Honor Win RT Launch Date, Colourways, RAM and Storage Configurations Revealed