BS Detector

AI-powered content credibility analysis — surface patterns, not judgments

BS Detector — AI-powered content credibility analysis Stop and think before you believe. BS Detector analyzes how web articles present information, scoring presentation quality and flagging claim hazards separately — so you always know WHERE the problem is. TWO-AXIS ANALYSIS: • Presentation BS Score (0-100): Measures evidence weakness, context loss, certainty inflation, emotional pressure, and source transparency • Claim Hazard (none/low/moderate/high): Flags whether the claims themselves are disputed, conspiratorial, unverifiable, or polarizing — independent of how well the article is written BRING YOUR OWN AI: Works with your own API keys — no data sent to us, ever. • OpenRouter (access to 100+ models including Claude, Llama, Mistral) • Google Gemini • Grok (xAI) Free models available — start analyzing at zero cost. FEATURES: • One-click analysis from any web page • SVG score gauge with green-to-brown palette • Component score breakdown with visual bars • Red flags list • Main claims extraction • YouTube support (description + transcript analysis) • Works on news sites, blogs, forums, and video pages PRIVACY: • Your API keys stay in your browser (chrome.storage.local) • No telemetry, no tracking, no data collection • Content is sent directly from your browser to your chosen AI provider • We never see your browsing data or analysis results BS Detector is open source: https://github.com/rdavidescu/bs-detector Scores are estimates, not verdicts. Think for yourself — we just help you see the patterns.