AI Compare · Test kit
Change one thing in a picture and nothing else
A flat illustration of a desk with seven distinct objects. The task is to turn the red mug blue. Everything else must stay exactly where and how it was, so you can compare the two pictures side by side.
Prices for United States, before sales tax Data release 2026.10.16-3, Oct 16, 2026 Rules 2.1.0
Run the test
- Download the file for the AI.
- Copy the prompt.
Attach desk-scene.png and ask: Change the red mug to a blue mug. Change nothing else in the picture: the same objects in the same places, the same colours, the same size and shape of image. Return the edited image.
- Open the AI in another tab, in your own account, start a fresh chat, attach the file and paste the prompt.
- Note what you used: the plan, the visible model or mode, whether web search was on, how many attempts it took and anything that went wrong.
- Score the answer with the checks below. Run the same kit in each AI you are choosing between.
Scorecard
Critical checks are marked. One failed critical check matters more than a nice style.
- The mug is now blue and keeps its handle and steamCritical
- Every other object is still there and unchanged: framed picture, window with sun, three books on the shelf, lamp, notebook with its red band, plant in its potCritical
- Nothing moved: compare the two pictures side by sideCritical
- Nothing added: no text, logo, extra object or new styleCritical
- Same 4:3 shape, not cropped
- All other colours are the same as before
To compare without knowing which AI made which, ask someone else to save the pictures as A, B and C before you score them, and to tell you which is which afterwards.
Reference answer (open after you score)
The only correct change is the colour of the mug. Typical failures: a redrawn scene in a different style, a moved or resized object, a recoloured notebook band (it is also red), a lost steam line, a cropped or square image, or the whole picture regenerated so that small details shift.
Kit version 1.0.0, files frozen on Oct 13, 2026. Original fixture by What the tech, released under CC BY 4.0. Fictional people, companies and figures. A passing score shows how one answer did on this kit; it does not prove the tool is generally better.