ToxiBlock

Automatically filter toxic comments using Machine Learning (SentimentAI).

ToxiBlock is an AI-powered Chrome extension that automatically detects and blurs toxic comments while you browse. Powered by a custom machine learning model trained on the Jigsaw Toxic Comment Dataset, ToxiBlock analyzes comment text in real time — including automatic translation for non-English content — and blurs anything flagged as toxic. No manual filtering. No word lists. Just smarter browsing. ✦ How it works ToxiBlock scans comments as they load on the page, sends the text to a FastAPI backend running a trained toxicity classifier, and blurs harmful content before you even see it. Comments load dynamically? No problem — ToxiBlock uses MutationObserver to catch new content as you scroll. ✦ Features — AI toxicity detection powered by a model trained on 150k+ real comments — Automatic translation pipeline for multilingual support — Works on YouTube and Twitter/X — Instant on/off toggle via the extension popup — Lightweight and privacy-friendly: no data stored, no accounts required ✦ Why ToxiBlock? Word-blocklists break the moment someone changes a letter. ToxiBlock understands context — it was trained on real-world toxic language, not just a list of banned words. Toggle it on. Browse better.