We were constantly pausing podcasts and Substack posts to ask, “Wait, is that actually true?”. After too many 30-minute Google rabbit holes, we decided to automate the work.

What it does

• Extracts claims from articles, blogs, YouTube videos (and even images) • Assigns a verdict + confidence score using live web search, a 200 M-paper research index, and a few proprietary datasets • Flags likely deepfakes and manipulated images • Runs entirely in your browser — only the claim snippets leave your machine for scoring

Status

• Free during early access (no wait-list) • Read more here (factifi.me/blog)

Looking for feedback on

• Accuracy thresholds (where should we draw “unverified” vs. “false”?) • UX friction — what feels clunky? • Ideas for new features

Happy to answer anything in the comments—thanks for taking a look!