TruthLens AI Video Deepfake Detector is a free online tool that runs AI forensic analysis on YouTube videos to detect deepfakes, face swaps, and AI-generated synthetic footage. It returns a confidence score and a breakdown of the visual, temporal, and metadata signals it inspected.
AI Video Content Detector New
Is this video real, or AI-generated?
Paste a YouTube link for deepfake detection and AI video forensics. Spot synthetic faces and AI-generated content.
Guests: up to 30 min · Free account: up to 60 min
AI-based visual pattern detection — not 100% accurate. Use results as supportive indicators, not definitive proof.
Looking for credibility scoring?Video Legitimacy Checker
AI Video Detector FAQ
Deepfake detection — common questions
Direct answers about how the AI Video Deepfake Detector works, what it can spot, and how to use the confidence score.
TruthLens combines multiple forensic signals — face-region inconsistencies, temporal flicker, compression artifacts, and metadata anomalies — to produce a probability score rather than a binary verdict. Like every deepfake detector, no model is 100% accurate, so the score should be read together with the highlighted red flags and used as one input in a broader verification workflow.
Yes. You can analyze public YouTube videos for free without creating an account. Paid plans unlock longer videos, higher daily quotas, batch analysis, and downloadable PDF reports for teams and researchers.
Yes. The detector is trained on synthetic content from modern generative video models including OpenAI Sora, Google Veo, Runway Gen-3, Kling, and Pika, in addition to traditional face-swap deepfakes. New models are continuously added as they are released.
TruthLens does not download or rehost the source video. It analyzes the public YouTube stream and stores only the resulting report (score, flags, transcript snippets) under your account. You can delete any analysis from your workspace at any time.
Generic AI text detectors only inspect written content. TruthLens is purpose-built for video: it inspects frame-level visual artifacts, audio-visual sync, transcript credibility, and source signals together — the combination that matters for deepfake detection and video authenticity verification.