Factualizer

■ 🎯 Read the News Factually, Side by Side • Factualizer analyzes the news article you're reading and shows the result right next to the original, in a resizable side-by-side panel. • One click gives you a factual, style-preserving rewrite of the article, a neutral title and summary, and a transparent bias & truthfulness scorecard — with the exact quotes that earned every score. ■ ✨ Key Features • Factual rewrite that keeps the original's voice, structure, and images — with inline remarks on omissions, loaded phrasing, and out-of-context quotes • Academic scorecard: 19 rubrics across two separate axes — Truthfulness & Reliability and Bias & Framing — never blended into one number • Every score below perfect must cite verbatim evidence from the article • Optional web-research pass verifies claims and restores omitted context • Bring your own AI: connect the API key of the cloud AI service you already use, run fully on-device with the browser's built-in model (no key needed), or point it at a self-hosted local model — the full list of supported services is on the project page linked below • Optional: get offered analysis automatically on news articles, or auto-run on sites you choose • Results are cached locally for 30 days — revisiting an article is instant and free • Interface localized into 40 languages, following the article's language ■ 👥 Who It's For Factualizer is for news readers, researchers, students, journalists, and media-literacy educators who want to see what an article says once the framing, omissions, and loaded language are stripped away — and an auditable account of how it was scored. ■ 🔒 Privacy Factualizer has no server. Article text is sent only to the AI service you configure, using your own API key — never to us or anyone else. Your key is stored locally in your browser and used only to call the service you chose. Analyses are cached locally and expire after 30 days. With the browser's built-in on-device model, nothing leaves your browser at all. ■ 🔗 Project Page & Feedback https://github.com/mjlee1983/factualizer/