Developer API
Detect AI-generated images, NSFW content, violence and document tampering from your own application. Send an image and get structured JSON back in seconds.
Included on Every Plan
API access is included on every plan, including Free, so you can be up and running with an API key today, no credit card needed. Each request uses one credit from your plan's balance, with 5 credits a day on Free. Document tampering checks and heatmaps start on Starter.
# Example request
curl -X POST https://sightova.com/api/v1/detect/ai \
-H "X-API-Key: YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{"image_url": "https://example.com/image.jpg"}'
A Simple REST API
Call it from any HTTP client. Send an image, get probabilities and the evidence behind them.
Generate API Key
Create a free account, then open the API Keys section in your dashboard to generate a key. Send it in the X-API-Key header, or as a Bearer token.
Send Request
Send a POST request to an endpoint under /api/v1/detect. Pass a publicly accessible image_url, an image_base64 string, or upload the file as multipart/form-data. One image per request, up to 50 MB.
Parse JSON
Receive structured JSON with the AI probability, a confidence value and a verdict, plus C2PA and watermark evidence when present. NSFW, violence and document tampering each have their own endpoint.
Clear, Actionable Data
The API returns the likelihood that the image is synthetic, a verdict, and the supporting evidence. NSFW, violence and document tampering are separate endpoints, each with its own result.
- "ai_probability" float (0.0 to 1.0)
- "verdict": ai-generated, real or uncertain
- "request_id" for tracking each request
"request_id": "req_5oQjGqU4mB2nQp9Hg7vTz",
"endpoint": "/api/v1/detect/ai",
"result": {
"ai_probability": 0.94,
"real_probability": 0.06,
"confidence": 0.88,
"verdict": "ai-generated"
}
}
Published Test Results
Check our work before you write a line of code. We test Sightova on 250k+ benchmark images and 50+ generative models, and publish the method and the results on our testing methodology page.
- Scores You Control: The API returns an AI probability (0.0 to 1.0), a confidence value and a verdict. You set the threshold and decide what happens next.
- Evidence for Your Reviewers: Invisible Google SynthID and OpenAI watermarks and C2PA credentials can name the exact tool, so your team sees what a result rests on.
- Built for Images in the Wild: The model reads the pixels, so detection keeps working when metadata has been stripped.
Ready to integrate?
Create a free account, generate an API key and send your first image today, no credit card needed. Upgrade when you need a larger monthly credit balance.