Veritype — Bias & Counterarguments

See the political framing in what you're reading, and the strongest case for the other side — in a panel beside the page.

Every article has an angle. Veritype shows you where. Not because the writer is lying — usually they aren't. Framing is quieter than that. It's which word gets used for the same event, which sources get quoted and which don't, what's stated as settled and what's left out entirely. It's the hardest thing to notice in a piece you already agree with. Veritype reads the page you're on the way an editor would, and puts what it finds in a panel beside the article. Where the framing sits on the political spectrum. The specific choices that put it there. And the strongest case against it — not a strawman, the real one. HOW IT WORKS 1. Read anything. Any article, blog post or op-ed. No setup, no separate app, nothing to paste anywhere. 2. One keystroke, or one right-click. Press Alt+Shift+V — or set your own key — to analyze the page you're reading. Or right-click and choose "Analyze with Veritype" on the whole page, or on just the passage you selected. 3. Decide for yourself. Results appear in seconds in a side panel, next to the text they describe. WHAT YOU SEE IN THE PANEL • A position on the spectrum, with the specific signals behind it — the actual phrases and choices, quoted, not just a label and a colour. • Counterarguments. The strongest opposing case, stated properly. This is the part people tell us they didn't expect and keep coming back for. • AI-content signals. Hedged observations about whether the writing shows patterns common in AI-generated text. Never a verdict — see the limits section below, we mean it. • Similar Sources (Pro). Drag a slider anywhere on the spectrum and Veritype finds real, current coverage of the same story from outlets at that position. Not a list of publications — actual articles about the thing you're reading right now. The outlets it draws on are US publications; the bias analysis itself works on any English-language article, wherever it was published. • A history of your recent analyses, so switching tabs doesn't lose the report you were reading. • Copy the whole report out with one click — properly formatted in Word, Docs and Outlook, clean prose everywhere else. Useful if you write, teach, or argue in writing. WHY THE OUTLET RATINGS ARE DIFFERENT Most source ratings collapse two unrelated questions into one number, and that's why people stop trusting them. An outlet can argue openly from the left or the right and report accurately. Another can claim neutrality and publish things that are simply false. So Veritype keeps them apart. Political lean is a judgment about framing. Factual record is a separate matter of documented history, and the two are shown separately, never averaged into a single score. A factual-record flag requires a specific, dated, publicly documented event — a correction, a retraction, a settled case — verified against primary reporting and cited on the spot. Never a third-party aggregator's rating, and never a general impression of an outlet's reputation. Where we couldn't confirm something to that standard, the outlet ships unflagged. Under-inclusion is the safe direction. The full method is published at veritype.app/sources, including the cases we decided NOT to flag and why. If you think a rating is wrong, that page is written so you can argue with it specifically. WHAT VERITYPE IS NOT It is not a fact-checker. It does not tell you whether a claim is true. It tells you how a piece is built and what the other side would say. It does not tell you what to think. There is no "trustworthy" badge, no score to defer to. The counterarguments exist so you can weigh them, not adopt them. The AI-detection signals are approximate and uncalibrated, and the panel says so where you read them. Treat them as a prompt to look closer, never as evidence. And it can be wrong. It's a model reading prose. The signals are shown so you can check its reasoning against the text rather than take its word. YOUR READING STAYS YOURS • Veritype never collects your browsing history. It only ever sees text you actively choose to submit. • Page URLs are never transmitted to our servers. • Your recent analyses are kept on your own machine and cleared when you close the browser. • No ads. No data sold. Not now, not later. • Prefer your text never touch our servers at all? Bring your own API key and it won't. FREE AND PRO Free gives you 10 analyses per rolling 24 hours — and articles someone else has already analyzed don't count against your limit, so popular stories are usually free to read. Pro is $5.99/month: effectively unlimited analyses, the specific AI-detection signals behind the summary, and Similar Sources. Cancel in one click, from inside the extension. WHO IT'S FOR People who read the news and would rather not be quietly steered by it. Students learning to read critically. Journalists checking their own framing before someone else does. Anyone who has finished an article, agreed with all of it, and wondered what that means. Every article has an angle. Now you can see where.