Google’s AI watermark

SynthID detector: what’s possible

SynthID is Google’s invisible watermark for AI images, video, audio and text. It can only be read by Google. GPTTrace can’t detect it — but it can read the labels Google puts next to it, and check the content itself.

Free · no sign-up · your file never leaves your device

Drop an image, click to choose, or paste

JPEG · PNG · WebP · AVIF · HEIC (Safari) — checked on your device

SynthID in one paragraph

Google DeepMind introduced SynthID in 2023 to watermark images from its Imagen model, and has since extended it to video (Veo), music and speech (Lyria and other audio models) and text (Gemini). The watermark is embedded by a neural network during generation and is invisible to people. A matching detector network can later recognise it, even after the content has been cropped, compressed or re-recorded. The detector is the part Google keeps to itself.

What GPTTrace checks instead

Google rarely relies on SynthID alone. Its products also write standard metadata, and that metadata is open to everyone:

  • IPTC digital source type. Google labels AI images with trainedAlgorithmicMedia, which many apps display as “Made with Google AI”. GPTTrace reads it and treats it as proof.
  • Content Credentials. As a C2PA steering member, Google has been adding signed manifests to content from its products. GPTTrace reads the signer and the actions.
  • Visible marks. Consumer Gemini images and Veo videos have carried small visible watermarks; a suspiciously cropped corner is a hint.

When the metadata has been stripped, GPTTrace’s neural classifier and forensic checks examine the content itself, independent of any watermark. Those checks work on images from any generator, not only Google’s.

A practical approach

If you suspect an image, clip or song came from a Google model, run it through GPTTrace first: if the IPTC label or a Google-signed credential is present, you have your answer in seconds without uploading anything. If not, use Google’s own SynthID verification for a definitive yes or no on Google content, and treat the GPTTrace classifier result as the general-purpose check for everything else.

The bigger lesson is that watermarks and labels only cover the companies that use them. Open-source models running on someone’s own computer add neither, which is why content analysis remains necessary.

Why this page doesn’t pretend

Search for “SynthID detector” and you will find sites that claim to detect it. They can’t, unless they are Google. We would rather explain what is actually possible and give you the checks that do work, than show a fake “SynthID: not found” that would mislead you into trusting an image.

What Google Imagen / Gemini and Google Veo and Google Lyria / MusicFX leave in a file

These are the traces GPTTrace checks for. “Some files only” means the trace is often missing — a re-save, screenshot or social-media upload removes metadata — so its absence proves nothing.

GeneratorWhereWhat to look forHow reliable
Google Imagen / Gemini (Nano Banana)PixelsSynthID invisible watermark — only Google’s own tools can read itDocumented by the vendor
Google Imagen / Gemini (Nano Banana)IPTC / XMPDigitalSourceType = trainedAlgorithmicMedia (“Made with Google AI”)Documented by the vendor
Google Imagen / Gemini (Nano Banana)C2PA manifestContent Credentials on images from recent Gemini and Pixel versionsSome files only
Google Imagen / Gemini (Nano Banana)PixelsA small visible sparkle watermark in the corner on free-tier Gemini imagesCommonly seen
Google VeoFramesSynthID watermark in every frame (not readable outside Google)Documented by the vendor
Google VeoFramesVisible “Veo” watermark on most consumer exportsCommonly seen
Google VeoC2PA / XMPContent Credentials or IPTC DigitalSourceType on some exportsSome files only
Google Lyria / MusicFXAudioSynthID audio watermark — only Google’s tools can read itDocumented by the vendor

How accurate is it? Our measured numbers

We test GPTTrace on labelled image samples and publish the results, including where it does badly. It is tuned to keep false accusations rare, so it misses some AI content rather than flag real work.

image check: AUC 0.938 (cross-validated)

679 labelled samples (399 AI, 280 human), run 2026-10-08. At the “Likely AI” line it caught 64% of AI samples and wrongly flagged 5% of human ones.

SourceTruthSamplesResult at “Likely AI”
gemini-nano-bananaAI4040% caught
midjourney-v6AI4068% caught
midjourney-v5AI4073% caught
flux-devAI4013% caught
flux-schnellAI4048% caught
sdxlAI40100% caught
gpt-imageAI4030% caught
klingAI3997% caught
leonardo-stablecogAI4098% caught
bitmind-imagine-mixAI4080% caught
fullsize-photosHuman4010% wrongly flagged
open-images-photosHuman405% wrongly flagged
lfw-facesHuman400% wrongly flagged
caltech-objectsHuman403% wrongly flagged
coco-photosHuman400% wrongly flagged
ffhq-facesHuman400% wrongly flagged
celeba-facesHuman4018% wrongly flagged

Data sources and method: methodology & accuracy.

Frequently asked questions

Why can’t other tools detect SynthID?
SynthID works like a secret key: the detector needs the same model Google used to embed the watermark. Google has not released that detector publicly, partly so that the watermark can’t be studied and removed more easily. Any third-party site claiming to read SynthID is, at best, reading metadata.
How can I check for SynthID?
Use Google’s own tools. Google has offered a SynthID Detector portal and built verification into some of its products, including asking Gemini whether an image was made with Google AI. Availability has varied by country and account type.
What does SynthID survive?
Google designs it to survive common changes: cropping, resizing, colour adjustment, compression and screenshots for images, and re-encoding for audio and video. Heavy edits and some deliberate attacks can weaken it.
Does SynthID cover images from other companies?
No. SynthID marks content from Google’s own models — Imagen, Gemini, Veo, Lyria and Google’s text models. An image from Midjourney or ChatGPT has no SynthID, so a negative SynthID check says nothing about other generators.
Is SynthID used for text?
Google applies SynthID to some text generated by Gemini, by subtly biasing word choices. It has published the text-watermarking method as open source, but detecting it still requires the specific key used for Gemini, so it cannot be checked from outside.