Last updated: 26 May 2026

<aside> 🛡️

HateBlocker for Chrome · On-device AI hides toxic comments on YouTube, Facebook, X and Instagram. Nothing leaves your browser. Free, no account required.

</aside>

A calmer comment section.

HateBlocker uses Chrome's built-in Gemini Nano model to classify comments locally — no data ever leaves your browser. Toxic comments are hidden automatically as you scroll. You stay in control.

• 100% local — no servers, no accounts
• Free in v1
• Multilingual

<aside>

</aside>

Features

<aside>

</aside>

On-device · No data leaves your browser

A calmer comment section. Powered by Chrome's built-in AI.

HateBlocker hides toxic comments on YouTube, Facebook, X and Instagram using Chrome's local Gemini Nano model. You stay in control.

• Comments are classified entirely on your device
• Toxic comments replaced with a discreet placeholder
• Click any hidden comment to reveal it instantly

Audit mode

See what's being flagged — without losing the comment.

Outline mode keeps the comment visible with a flag badge. Click the badge to hide it. Great for tuning sensitivity or reviewing edge cases.

<aside>

</aside>

<aside>

</aside>

Total control

Filter exactly what you want to filter.

Pick a sensitivity level, choose which categories to hide, and decide how flagged content should appear. Toggle platforms independently. Every change applies live — no page refresh.

• 3 sensitivity levels
• 6 category filters
• 3 hide styles (subtle, prominent, outline)
• Per-platform toggles
• Optional on-device translation

Multilingual · On-device translation

Catches hate in Norwegian, Spanish, German, and more.

Toggle on translation and HateBlocker detects each comment's language with Chrome's on-device Translator API before classifying. All locally. No cloud round-trip.

<aside>

</aside>

Summary

HateBlocker is a Chrome extension that hides toxic comments on YouTube, Facebook, X (Twitter) and Instagram using Chrome's built-in on-device AI (Gemini Nano). No comment text, no browsing history, and no personal data is ever sent to a server. All analysis happens locally inside your browser.

What HateBlocker does

When you visit YouTube, Facebook, X (Twitter) or Instagram, HateBlocker reads the visible comment elements on the page, sends the text to Chrome's local Gemini Nano model, and hides comments that the model classifies as toxic under the categories you have selected.

Data we process

• Comment text on the pages you visit — used only for local classification by Chrome's on-device AI. Not stored. Never transmitted.
• Your settings (sensitivity, hide style, filter categories, per-platform toggles, translation toggle) — stored in chrome.storage.local on your device only. Never transmitted.
• Count of comments hidden — stored locally to display in the popup. Never transmitted.

What HateBlocker does NOT do

• It does not send any data to any remote server.