Upload an image, get a verdict. Where the file carries signed content credentials we verify them cryptographically — proof, not a probability. Where it does not, the detection model scores the pixels.
1,000 free credits · no credit card · PNG, JPG, GIF, WebP up to 10MB
/api/v1/detect/imagecurl -X POST https://www.wasitaigenerated.com/api/v1/detect/image \
-H "Authorization: Bearer YOUR_API_KEY" \
-F "file=@/path/to/image.jpg"Content-Type — multipart/form-data
file — required, PNG · JPG · JPEG · GIF · WebP
Max size — 10MB per request
Cost — 300 credits per image
{
"isAI": true,
"confidence": 0.99,
"verified": true,
"c2pa": {
"vendor": "OpenAI",
"generator": "ChatGPT",
"issuer": "OpenAI",
"signedAt": "2024-08-06T21:53:37+00:00",
"signatureValid": true,
"sourceType": "trainedAlgorithmicMedia",
"isAI": true
},
"patterns": [
"ai_verified",
"c2pa_signed",
"synthetic_artifacts"
],
"analysis": {
"likelihood": "high",
"reasoning": "AI-generated imagery detected with 99.0%
confidence. Verified AI-generated: content
credentials signed by ChatGPT (OpenAI) on 2024-08-06."
},
"detailed": {
"scores": { "ai": 99.0, "human": 1.0 },
"model": "v1.1"
}
}Most image detectors give you a probability and leave you to defend it. This endpoint tells you when it has actual proof.
Read the evidence behind an image
Workflow illustrationModel analysis
An estimate from the image itself. Read the confidence alongside the AI/human assessment.
Content credentials
When present, inspect the C2PA signature and declared source. Check what the credential actually establishes.
Generators that follow the C2PA standard sign their output with content credentials. When those credentials are intact, we validate the signature — the image is AI-generated, and that is not a judgement call.
Use it when you need a result you can put in front of a customer, a moderator or a regulator.
Most images in the wild have been screenshotted, re-encoded or stripped of metadata. For those, the detection model scores the pixels and returns a confidence plus an AI/human split.
Strong evidence — but set a threshold and keep a human in the loop at the boundary.
Multipart upload, Bearer auth, JSON back
curl -X POST https://www.wasitaigenerated.com/api/v1/detect/image \
-H "Authorization: Bearer YOUR_API_KEY" \
-F "file=@/path/to/image.jpg"isAI — boolean verdict.confidence — 0–1 from the detection model.verified — true when signed content credentials prove AI generation.patterns[] — machine-readable tags such as ai_verified or synthetic_artifacts.detailed.scores — AI vs human probability as percentages.Check damage photos and receipts at intake, before an adjuster spends time on a claim built on generated evidence. See the use case
Score every user upload as it arrives and route synthetic media into your existing moderation queue automatically.
Catch generated profile pictures at signup, where fake accounts are cheapest to stop. See the use case
Flag synthetic ID and document images during identity verification, alongside your existing liveness checks.
Verify user-submitted and wire imagery before publication, and log the result as part of your editorial audit trail.
Detect generated product photos that misrepresent what a seller actually ships.
Every failure is a JSON body with an error field
300 credits per image — no subscription needed to start.
1,000
credits on email verification
3 images to try the endpoint. No card required.
$5 – $29
one-time, never expires
25,000 credits ≈ 80 images. 200,000 credits ≈ 660 images.
$299/mo
no per-image cost
Unlimited calls, whitelabel PDF reports, bulk processing.
A JSON object with isAI (boolean), confidence (0–1), verified (boolean), c2pa (the parsed content credential or null), a short analysis, and detailed.scores giving the AI and human probabilities as percentages. The verified flag is the important one: when it is true, the image carries a validly signed content credential declaring it AI-generated, so the result is a fact rather than a probability — and c2pa tells you who signed it (vendor and issuer, e.g. OpenAI), which generator (e.g. ChatGPT) and when.
confidence is the detection model’s probability that the image is synthetic — useful, but statistical. verified means the image carries a C2PA content credential whose signature verifies and which declares the asset AI-generated, which proves it. The full manifest is returned under c2pa: issuer is the certificate issuer and cannot be spoofed, signatureValid is the cryptographic check, and sourceType tells you whether the credential claims AI generation (trainedAlgorithmicMedia) or a camera capture (digitalCapture). Treat verified: true as certain, and confidence as strong evidence that still deserves a human decision at the boundary.
Images from the current generation of diffusion and transformer image models — Midjourney, DALL·E, Stable Diffusion, Flux, Firefly, Imagen and the tools built on them. Generators that embed content credentials are additionally caught by cryptographic verification, independent of what the visual model sees.
PNG, JPG, JPEG, GIF and WebP, up to 10MB per request. Send the image as multipart/form-data with the field name file. Very large images may still hit an upstream size limit — downscale to a few megabytes for the most reliable throughput.
300 credits per image. A new verified account starts with 1,000 free credits, which covers three test images. From there, the $5 pack of 25,000 credits is roughly 80 images, the $29 pack of 200,000 credits is roughly 660, and the $299/month plan removes the per-image cost entirely.
Heavy compression, cropping and screenshotting all remove signal, and screenshots strip content credentials entirely — which is why a screenshot of an AI image can come back with verified: false and c2pa: null. Copying an image to the clipboard has the same effect. Send the original file whenever you have it, and treat a re-encoded copy as weaker evidence.
Yes. Call the endpoint on upload, store the verdict alongside the asset, and route anything above your confidence threshold into an existing moderation queue. For sustained high volume, the unlimited plan removes per-image cost; contact us if you need an SLA or dedicated limits.
Images sent through the API are processed to produce a result and are not used to train our models. See the privacy policy for the full detail on retention.
Create an account, verify your email, and your API key is waiting in the dashboard.