DeepfakeDetector.ai API: Image & Video Verification Guide

If your platform accepts user-uploaded images or videos, synthetic media is already infiltrating your pipeline. Profile photos, onboarding selfies, and user-generated content clips submitted to newsrooms all contain a growing fraction of generated or manipulated material that requires active moderation.

This guide covers adding detection using the DeepfakeDetector.ai API: acquiring an API key, scanning images, handling asynchronous video processing, defensively parsing responses, and designing your product around probabilities rather than absolute facts. All endpoints reside behind a single base URL at https://app.deepfakedetector.ai/api/v1.

For images, the API targets AI-generated visuals and faces from major generators like Midjourney, DALL·E, Stable Diffusion, and Flux. For videos, it covers face swaps, lip-sync manipulation, and fully AI-generated clips analyzed frame by frame. API access begins on the Starter plan, while the free tier provides 50 monthly detections via the web app for evaluation purposes.

Once you subscribe to a plan with API access, your secret key is displayed only once. You must copy it immediately into your secrets manager and keep it strictly server-side. Requests require a Bearer token, and exposing a key within a browser bundle poses severe security risks.

Your first image request can be made by sending a publicly reachable HTTPS URL to the detect/image endpoint. Optional parameters allow the use of short-lived signed URLs for private storage files. Alternatively, if you possess the file locally, you can submit it as multipart form data supporting MP4, MOV, WEBM for video, and JPG, PNG, WEBP for images.