Paste any link — YouTube, news, X post — get a credibility report in seconds. No signup, no wait.
In the past 12 months, AI-generated articles and deepfake videos have flooded every platform. A single viral fake can sway public opinion, tank a stock, or ruin a reputation. On Hacker News, a thread about “Bullshit Detector” racked up 63 comments in hours — developers are terrified of the misinformation wave, and existing tools simply can't keep up.
Manual fact-checking sites like Snopes are thorough but slow — they only cover the top 0.1% of viral claims. Meanwhile, automated solutions like NewsGuard rate entire domains, not individual pieces of content. That means a perfectly legitimate news site can publish a single false story, and you'd never know until it's too late.
We're at a tipping point: LLMs can now extract claims and cross-reference them with authoritative databases in under 3 seconds. The technology is ready, the demand is urgent, and the window to build trust back into the web is now. That's why we built Bullshit Detector — not as a developer API, but as a tool anyone can use.
Copy the URL of a YouTube video, news article, blog post, or X/Twitter thread. Our system automatically fetches the full text or transcript — no manual copying required.
We use a fine-tuned LLM to pull out 3–5 key factual statements from the content. Each claim is then compared against trusted sources like Reuters, AP, PubMed, and Snopes via real-time search.
Within seconds you'll see a score (0–100), a verdict (✅ Trusted / ⚠️ Suspicious / 🚩 False), and direct citations for every claim. Share the report or dig deeper — it's up to you.
No waiting for human fact-checkers. Paste a link and get a full credibility report in under 10 seconds — fast enough to check before you share.
Every claim is linked back to its original source — whether it's a peer-reviewed study, a government database, or a known debunk. You see exactly where the evidence comes from.
No need to read a wall of text. The report uses a clear color-coded score and badge — ✅ Trusted, ⚠️ Suspicious, or 🚩 False — so you can decide in a second.