Debunkr vs Glowby

Debunkr helps you read the news like a skeptic — without the homework. Paste any article link and Debunkr scores its credibility from 0 to 100 in seconds, then shows its work so you can judge for yourself. WHAT YOU GET INSTANTLY • A 0–100 credibility score with a plain-English rating: Highly credible, Mostly credible, Mixed, Questionable, or Low • A Rating Strength meter that tells you how much to trust the score • A breakdown across publisher reputation, article quality, and tone & balance • Devil's Advocate — the single strongest reason the score might be wrong GO DEEPER WITH DEEP ANALYSIS • A Skeptic and an Advocate AI debate the article point by point • A Judge weighs both sides into a final verdict • A fact-check pipeline pulls up to 3 checkable claims and runs them against the Google Fact Check Tools API (PolitiFact, Snopes, AFP, Reuters Fact Check, and more) • Named people, places, and organizations are cross-checked against Wikipedia • Every citation comes from a real search result — no invented sources HOW THE SCORE WORKS Debunkr blends three signals: • Publisher reputation (45%) — a hand-curated list built from Wikipedia's Perennial Sources consensus and the IFCN network of certified fact-checkers • Article quality (45%) — an AI model reads the piece and rates sourcing, hedging, and writing • Tone & balance (10%) — a small penalty for extreme political slant either way BUILT TO BE HONEST Debunkr rates credibility signals, not absolute truth. A polished hoax can still score Mixed, and breaking news often has no fact-checks yet — when that happens, Debunkr says so instead of guessing. AI can make mistakes; we add safeguards but errors are possible. The in-app Methodology and Limitations screens lay it all out. NO ACCOUNT, NO TRACKING No name, no email, no login. A random device ID enforces fair-use limits and nothing more. We don't sell or share your data. Debunkr is a media-literacy tool, not a verdict machine. Use it to ask better questions — then decide for yourself. Debunkr is in alpha. Ratings are estimates, not verified facts. Always check important claims independently.

The internet doesn't slow down for the truth. Glowby is the fast lane. Paste any YouTube Short, TikTok, news article, or typed claim — Glowby's AI watches the video, listens to what's said, reads what's on screen, hunts down evidence across the web, and returns a truth score from 0.0 to 9.9 with the sources laid out so you can judge for yourself. Ask a question instead, and Glowby gives you a straight, sourced answer. HOW IT WORKS The Ears transcribe what the video says The Eyes read what the video shows — on-screen text, charts, headlines The Evidence agent searches fact-check databases and the live web The Judge scores every claim against a written rubric, with sources The lowest-scoring central claim sets the headline score — one number, no cherry-picking WHAT MAKES GLOWBY DIFFERENT Shows its work: every verdict names its sources, every score has a written methodology you can read at glowby.io/about Honest about uncertainty: thin evidence is labeled thin — never dressed up as certainty A report button on every result — a real person reviews every report No account, no ads, no tracking profiles Glowby is a public beta. It can be wrong — that's why every claim ships with its sources and a one-tap way to tell us when we miss. Built by a 17-year-old who thinks checking a claim should be as easy as sharing it.