How Google labels Gemini images
Google marks images from Gemini, Nano Banana and Imagen in two ways: an invisible SynthID watermark woven into the pixels, and metadata in the file (IPTC “trainedAlgorithmicMedia” and, increasingly, Content Credentials). SynthID can only be read by Google’s own tools, but we read the metadata — so an original Gemini download is confirmed instantly. Files named “Gemini_Generated_Image_…” are recognised too.
For screenshots and re-shared copies, our pixel detectors take over.
Nano Banana edits are trickier
Nano Banana is often used to edit real photos — changing a background, an outfit or a face. Edited photos can look mostly real to any detector. If the file is labelled as AI-edited we tell you; otherwise look closely at the edges of people and objects, mismatched lighting and text in the background.
Frequently asked questions
Can you read Google’s SynthID watermark?
No — only Google can verify SynthID (you can ask Gemini itself whether an image has it). We read the file labels Google adds and run independent detectors on the pixels.
How accurate is it?
When a file still carries Content Credentials or other AI labels (common for ChatGPT, Gemini, Firefly and Stable Diffusion files) the answer is essentially certain. Otherwise we combine several open-source detectors; they are right most of the time on unedited AI images but can be fooled by screenshots, heavy compression, filters and brand-new generators. That is why every result shows a confidence level and the evidence behind it — treat it as strong evidence, not proof.
Is IsItMadeByAI really free?
Yes. Every scan is free, there is no sign-up, no credit card and no watermark on your results. The site runs on open-source detection models, which is how we keep it free.