Detect AI-generated faces and deepfakes via REST API. Server-side inference with EfficientNet-B4 + frequency analysis.
POST /api/v1/detectGET /api/v1/detectAll requests require a Bearer token in the Authorization header.
Authorization: Bearer dc_live_your_api_key_hereContact us to get an API key. Keys have read/write permissions and usage is tracked.
/api/v1/detectDetect AI-generated content| Field | Type | Required | Description |
|---|---|---|---|
image | string | Yes* | Base64-encoded image (JPEG, PNG, WebP) |
images | string[] | Yes* | Array of base64 images (batch mode, max 10) |
cropFace | boolean | No | Auto-crop face region (default: true) |
webhookUrl | string | No | URL for async result callback |
* Provide either "image" (single) or "images" (batch)
{
"success": true,
"data": {
"authenticityScore": 92, // 1-100 (100 = definitely real)
"verdict": "real", // real | suspicious | likely_fake | fake
"confidence": "high", // high | medium | low
"pFake": 0.0842, // Raw P(AI-generated) 0-1
"modelVersion": "v3-efficientnet-b4",
"processingMs": 245
}
}/api/v1/detectModel info + health check{
"success": true,
"data": {
"status": "ready",
"model": {
"modelVersion": "v3-efficientnet-b4",
"inputShape": [1, 3, 224, 224],
"fileSizeMB": 70.2,
"runtime": "onnxruntime-node"
}
}
}curl -X POST https://deep-check-two.vercel.app/api/v1/detect \
-H "Authorization: Bearer dc_live_YOUR_KEY" \
-H "Content-Type: application/json" \
-d '{
"image": "<base64-encoded-image>",
"cropFace": true
}'| Status | Meaning |
|---|---|
400 | Bad request — missing image or invalid format |
401 | Unauthorized — invalid or missing API key |
413 | Image too large — max 10MB |
429 | Rate limit exceeded |
500 | Server error — model loading or inference failure |
Deep-Check API by HIUM Solutions SL
Contact: info@deep-check.io | Andalucia, Spain