I've spent the past months building Rhetoriq, and I'm launching it here today. Most argument-analysis tools work fine on neat textbook examples. I wanted something for actual messy text: a Reddit comment, a heated reply, a headline, where something valid and something fallacious often sit in the same sentence. You paste, speak, or select text anywhere on your phone. That last one is the part I'm proudest of: select text in WhatsApp, a browser, an email, anything, and Rhetoriq shows up right in the selection menu. Tap it, and the analysis appears over your screen without opening the app or switching context. It flags the weak passage, names the fallacy, notes what's strong, and scores the whole argument 0–100. Useful for checking your own reply before you hit send, too. Example: "Solar is cheaper than coal, so anyone against it is just pushing oil propaganda." → strong point on the cost claim, ad hominem on the rest. It's built around a fixed set of 16 commonly used fallacies plus manipulation techniques like false balance, loaded language, and manufactured consensus. For each text it pulls out strong points, weak points, and any fallacies or rhetorical tricks it finds. Each tied to a specific passage rather than a generic label slapped on the whole thing. A false positive still happens from time to time. The goal was to beat generic "explain this argument" AI prompting, not to be infallible. No subscription, you buy credits and they don't expire. https://play.google.com/store/ap... (Web app for bigger screens: https://rhetoriq.nl/app?lang=en)

Rhetoriq

Fallacy detection for any real-world text

Fallacy detection for any real-world text