Developer docs for the Detect API
Detection is the active API track. Submit an image file, receive a verdict, confidence level, evidence summary, scan ID, and report URL.
Run image authenticity checks and return verdicts, evidence signals, reports, and confidence bands.
Same auth and response shape as image detection, plus a per-frame AI-probability breakdown.
Generate polished visuals from prompts once Pixoraid Ink AI goes live.
Automate PDF, document, spreadsheet, and writing tasks from your own tools.
Ship code generation, debugging, and explanation workflows from one endpoint.
POST /detect
x-api-key: <partner_api_key>
Content-Type: multipart/form-data
file=<image_file>
Response:
{
"scan_id": "scan_...",
"verdict": "ai" | "real" | "uncertain",
"confidence_level": "high" | "medium" | "low",
"recommended_action": "allow" | "review",
"evidence_summary": "...",
"report_url": "/scans/.../report"
}Video works the same way — POST /detect/video takes the same auth and file upload, plus a duration field, and returns the same verdict shape with a per-frame breakdown. See the AI video detector page for the public-site version.
Signed-in users can also call the API with their Google ID token as `Authorization: Bearer` — usage counts toward the account's monthly plan.