ChatGPT’s New Image Update: A "Creative Studio" Inside Your Chat
OpenAI's new GPT Image 1.5 update turns ChatGPT into a real design tool. Featuring a dedicated "Images" sidebar and 4x faster speeds, the biggest change is "granular editing," fixing specific parts of a photo without changing the rest.
A screenshot of the new ChatGPT "Images" interface showing the sidebar, style presets, and the new granular editing tools in action.
If you opened ChatGPT this week, you might have noticed a change in how it handles photos. Around mid-December (specifically December 17, 2025), OpenAI rolled out a significant update powered by a new model called GPT Image 1.5.
This isn't just "DALL-E 3 with a new name." It feels less like a chatbot trying to draw and more like a proper design tool built directly into the interface. Here is a breakdown of what has changed and why people on social media are talking about it.
What Changed?
The biggest shift is that image generation now has its own dedicated home. Instead of just typing prompts into the standard chat box, there is now an "Images" section in the sidebar.
Think of this as a "creative studio" mode. When you enter this mode, the interface changes to offer preset filters, aspect ratio toggles, and style suggestions, similar to tools like Canva or Photoshop, but powered entirely by AI prompts.
The Big Feature: "Granular" Editing
For a long time, the frustration with AI images was the "reroll" problem. You would generate a perfect image of a dog on a skateboard, but if you asked to "make the skateboard red," the AI would draw a completely different dog.
GPT Image 1.5 solves this with granular editing.
- Precision: You can now change specific parts of an image without ruining the rest. If you want to change a character's shirt from blue to green, the model updates just the shirt. The face, lighting, and background stay exactly the same.
- Text Handling: It is finally much better at spelling. Logos, signposts, and labels on products are coming out clearer and more readable.
Speed and Performance
The new model is reported to be 4x faster than the previous version. This effectively removes the long "thinking" pause that used to break your creative flow. You can now iterate, generating, tweaking, and regenerating almost in real-time.
What People Are Doing with It (Social Trends)
Since the rollout, a few trends have taken over social media feeds (X/Twitter and Threads):
- The "Ghibli" Portraits: Users are uploading their family photos and using the new style presets to turn them into Studio Ghibli-style anime characters. Because the new model keeps facial features consistent, the results look like you, not just a generic anime character.
- "Vibe Coding" for Designers: Web designers are sketching layouts on napkins, photographing them, and having the new model turn them into crisp, high-fidelity mockups in seconds.
- Ad Recreation: Marketers are taking vintage ads and asking the model to "modernize this for 2026" while keeping the exact same composition, which was previously very difficult.
Comparison to Rivals
This update comes right after a wave of competition from Google (specifically their "Nano" series updates). While Google has focused heavily on hyper-realism, ChatGPT’s update seems focused on control, giving you the tools to actually fix an image rather than just generating random pretty pictures.
How to Try It
- Open ChatGPT on the web or mobile.
- Look for the new "Images" icon in your sidebar (or the apps menu).
- Upload a photo to edit, or start fresh by clicking one of the style presets.
