IsWatermarked?

Google DeepMind

SynthID Checker

SynthID is Google's invisible watermark for AI images, video, audio and Gemini text. Important upfront: only Google can read SynthID. This page shows the official ways to check it — and what this site can verify locally instead.

How to actually check for SynthID

  1. Ask Gemini. Upload the image, video or audio to the Gemini app and ask “Is this made with AI?” — Google’s deployed consumer verification path, rolling out to Search and Chrome as well.
  2. SynthID Detector portal. Google’s dedicated verification portal exists but is limited to an early-tester waitlist (journalists and media professionals first).
  3. Enterprise API. Google Cloud offers an AI content detection API on its enterprise agent platform — not a consumer tool.

There is no public API this site could legitimately use to detect SynthID, so we don’t simulate one. If that changes, a clearly labeled official check will be added here.

What you can verify here instead: C2PA

SynthID (invisible pixels/audio watermark) and C2PA Content Credentials (signed metadata) are different layers, and Google now attaches or verifies both across its products. Many AI images — including every ChatGPT image since May 2026 and Pixel 10 camera photos — carry C2PA credentials that can be validated right here in your browser:

Your image is inspected locally in this browser tab. Nothing is uploaded.

SynthID vs C2PA in one minute

SynthIDC2PA Content Credentials
What it isInvisible statistical watermark in the content itselfCryptographically signed metadata attached to the file
Who can verifyGoogle only (Gemini, Detector portal, enterprise API)Anyone, with any conforming tool — including this site
Survives edits?Designed to survive moderate editsBreaks or is stripped by editing/screenshots — by design detectable
CoversImages, video, audio, Gemini textImages, video, audio, documents

Related