API Reference
Interactive reference for the Raid AI detection and fact-checking API, with a live Try-It for every endpoint.
The Raid AI API is a REST API for detecting AI-generated and manipulated images and audio, checking documents for tampering, and fact-checking media against the public record. Browse endpoints by modality in the sidebar — each page documents the method, parameters, request body, response schema, and status codes, with copy-paste samples in Python, cURL, JavaScript, PHP, Go, and Java.
Authentication
Every request is authenticated with a bearer token in the Authorization header. See
Authentication for details.
Synchronous vs. asynchronous
- Image, Audio, and Document endpoints are synchronous — the response contains the
result. Document analysis holds the request open until the verdict is ready (typically 90 to 120
seconds); a
202means it is still processing and can be re-fetched. - Fact-Checking endpoints are asynchronous — you submit a job, receive a
jobId, then pollGET .../jobs/{id}untilstatusisCompleted.
Try it
Every endpoint has a live Try it panel. Requests are sent through this site's secure proxy, so you can run real calls from the browser.
Use a real API key
Paste an API key from your dashboard into the
Authorization field to run real calls. The key must include the scope for the endpoint
you're calling (e.g. an image key for the image endpoints), or the request returns
403 Forbidden. The unauthenticated
health check needs no token.