Verify breaking news clips, political speeches, and viral videos on X. Paste any tweet URL or mention @SealedRoseScan in any reply for instant frame-by-frame forensic analysis.
Paste a link to any tweet containing video. Our forensic engine extracts keyframes directly from the post, inspects neural artifact signatures, and produces a cryptographic certificate.
Two Ways to Verify
Use our web scanner above or tag our automated bot directly inside the reply thread.
Encounter a breaking news clip, controversial speech, or sudden celebrity endorsement that looks artificially generated or out of sync.
Reply to the tweet mentioning our verified bot handle. Click below to copy the prompt or launch X directly:
The engine resolves video frames, scans temporal consistency, and issues an immutable cryptographic certificate with SHA-256 hashes.
Is this press conference video authentic or AI lip-synced? @SealedRoseScan verify video
@open_intel Forensic Verdict: Authentic Video (84.4% confidence). Frame temporal consistency confirms optical sensor capture with natural motion blur and facial dynamics.
Forensic Capabilities
Every video submitted via X link or direct upload undergoes deep frame extraction and optical coherence tests.
Detects inter-frame flickering, unnatural motion vectors, and micro-blur artifacts typical of generative video models (Kling, Sora, Runway Gen-3).
Scans mouth geometry and audio-visual synchronization to identify voice-driven avatar generators (HeyGen, SadTalker, Wav2Lip).
Genuine camera sensors impart unique photo-response non-uniformity. Synthetic videos display sterile pixel distributions lacking natural optical noise.
Explore specialized forensic tools for photos, videos, and cloned audio.
Frequently Asked Questions
You have two instant options: (1) Paste the tweet URL directly into the scanner above to inspect video frames, or (2) Reply to the video tweet on X and mention @SealedRoseScan (e.g. "@SealedRoseScan check video"). Our automated forensic engine will inspect the video keyframes and reply with an official forensic verification verdict.
Our neural inspection engine detects generative video architectures including Kling 1.5, Runway Gen-3 Alpha, Luma Dream Machine, OpenAI Sora, Wan 2.1, ByteDance Seedance / Dreamina, HeyGen, DeepFaceLab, FaceSwap, and synthetic audio lip-sync generators.
Video deepfakes frequently leave micro-artifacts across sequential frames: temporal jitter, boundary warping around jawlines, flickering teeth, pupil inconsistency, and optical sensor noise discontinuities. Sealed Rose extracts keyframes, computes pixel-level frequency spectra, and checks temporal motion coherence.
Yes. Every completed scan generates a 1-click "Post Verdict to X" share button and an immutable cryptographic verification record. You can post the verdict directly in reply to the original tweet to provide empirical proof to the community.
Yes. Video verification is 100% free with no software installation required. You can scan videos via our web portal or directly through our automated X bot handle @SealedRoseScan.
Run free neural analysis on any video file or social link with zero registration and zero video retention.
