In-memory processing · No saving · Probabilistic results

AI Image Detector

Upload an image and get a probabilistic estimate. Useful for quick screening of AI-generated or heavily altered content.

Start analysisHow it worksHow to interpretFAQ
JPGPNGWEBP

Good for

  • Spot-checking images shared on social apps (re-uploads, screenshots).
  • Assessing suspicious “too-perfect” photos in chats or posts.
  • Quick triage before deeper verification (source, context, metadata).

Limitations (important)

  • Not definitive proof: false positives/negatives are possible.
  • Heavy compression and edits can change the score.
  • Cannot identify the AI model/platform or confirm identity.

How it works

The system outputs a score (estimate) and a confidence level. Results can change with quality, compression, resizing, screenshots, and editing.


1) Upload (in RAM)
2) Visual analysis + quality signals
3) Request ends → memory freed

Interpreting the result

High score usually means stronger AI-like patterns. “Uncertain” means mixed signals: avoid strong conclusions.


Quick FAQ

Is it 100% accurate?
No. Treat it as an indicator, not proof.
What does “uncertain” mean?
Mixed/weak signals (borderline zone).
Do you store images?
Processing is designed to be in-memory. See the Privacy page for details.
No detector is infallible: use the result as a signal, not as proof.
Elaborazione in memoria · Nessun salvataggio immagini · Risultati probabilistici
AI Image Detector – check if an image is AI-generated – AI Image Detector