PROCESS
Upload anything. Every claim is indexed. Every contradiction is surfaced. Every receipt is exportable.
Upload video, audio, or text files directly. Paste a URL from YouTube, TikTok, Instagram, Facebook, Twitch, or any other platform. Record from your microphone for live analysis. Drag and drop multiple files at once. The platform accepts MP4, MP3, WAV, PDF, TXT, and dozens of other formats.
Supported sources: YouTube, TikTok, Instagram, Facebook, Twitch, Pinterest, Spotify podcasts, Apple Podcasts, direct file upload, microphone recording, live stream URLs.
Every claim, assertion, opinion, promise, and prediction is extracted from your media with full attribution. Each statement is linked to a speaker, timestamped to the exact moment it was said, and categorized by type. Speakers are identified by voice and matched across all your uploads — so one person’s statements from a 2019 interview are linked to their statements from a 2024 podcast.
Statement types: fact, opinion, prediction, promise, denial, position change.
The system compares every new statement against everything the same speaker has said across all your uploaded media. When positions conflict — when someone said one thing in January and the opposite in June — both statements are presented side by side with full timestamps, source links, and a severity ranking. You are notified immediately.
Contradiction severity: direct reversal, qualified shift, contextual inconsistency.
Export your findings in any format. Generate side-by-side video comparisons that play both contradicting statements back to back. Create PDF reports with timestamped source citations. Export social-ready clips formatted for Twitter, Instagram, TikTok, and LinkedIn. Share direct links that anyone can verify. Every output includes the original source references.
Export formats: video clips, PDF reports, SRT/VTT transcripts, social clips, shareable web links, raw JSON via API.
Real-time mode lets you process live content as it happens. Connect a live stream URL or record directly from your microphone. Statements are extracted and cross-referenced in near real-time — so contradictions can be flagged during a live press conference, debate, or interview rather than hours or days after.
As the speaker talks, each new statement is compared against their entire indexed history. When a conflict is detected, you see it immediately with the original contradicting statement, its source, and its timestamp. Available on Professional and Enterprise plans.