AI image detector

One AI score is not enough.

AI Screener compares whole-image and regional model signals with residual statistics, pixel structure, metadata, and provenance. It explains disagreement instead of hiding it behind a binary label.

01

Whole-image classifier

A pinned open model checks the complete scene for synthetic-image patterns.

02

Regional patch comparison

Overlapping regions help reveal composites where only part of the image may be generated.

03

Residual statistics

An independent research detector looks at signal residuals rather than visible subject matter.

04

Provenance and pixels

C2PA, metadata, noise, frequency, and recompression results provide independent context.

No automated detector can prove authenticity. AI Screener combines independent signals, shows the evidence, and leaves the decision with a human.

Run the free checker

Questions to ask of the result

Limits matter as much as the signals.

Can an AI image detector prove a photo is real?

No. A detector can report signals and limitations, but a negative result does not verify camera capture or human origin.