👉 Driectly download.
FB Feed Diet is a browser extension for Facebook designed to hide sponsored posts, recommended content, Reels, and ads from your news feed, decluttering your reading experience.
Folded posts are replaced with a sleek title bar that can be expanded anytime with a single click, without abruptly removing blocks or breaking page layout. You can also customize folding rules and display preferences to suit your reading habits.
| Before Filtering | After Filtering (Ads & Suggestions) |
|---|---|
 |
 |
- Block Ads (Sponsored Content): Folds sponsored posts, Marketplace recommendations, and search ads.
- Block Suggestions (Recommended Posts): Folds algorithmic recommendations like "Suggested for you".
- Hide Short Videos (Reels & Stories): Optional toggles to fold Reels and Stories carousels.
- Live Stats: Counts daily blocked ads, suggestions, and video items.
- Hover Preview: Hover your cursor over the title bar to preview full post content without clicking "See more".
- Expand & Collapse Anytime: Replaces folded content with title bars; click anywhere on the bar to reveal the post.
| Regular Post | Hover Tooltip Preview | Folded State |
|---|---|---|
 |
 |
 |
- Custom Folding Rules: Ideal if you prefer skimming headlines and clicking to expand only what interests you.
- Appearance Customization: Customize title bar height (18px or 36px, default 18px).
Coming soon.
- Go to the Releases page and download the latest Chrome ZIP package.
- Extract the ZIP file to a local folder and keep the folder intact for your browser to load.
- Open Google Chrome and navigate to
chrome://extensions/. - Enable Developer mode in the top-right corner.
- Click Load unpacked.
- Select the extracted project folder.
- Open Facebook.
Once loaded, the extension will be active on Facebook pages.
Currently loadable via Firefox's Temporary Add-on feature:
- Download or clone this repository.
- Open Firefox and navigate to
about:debugging#/setup/runtime/this-firefox. - Click Load Temporary Add-on….
- Select
manifest.jsonin the project directory. - Open Facebook.
Note: Temporary add-ons in Firefox are unloaded when the browser restarts; repeat step 3 to reload.
Click the FB Feed Diet icon in your browser toolbar to instantly toggle filtering and view current diet statistics.
Open "Options" in the extension popup:
- Folding rules for the five main post categories.
- Title bar height, title snippet display, hover text preview, and more.
- Interface language (English / Traditional Chinese / Spanish) and Light / Dark theme mode.
- Detection engine (Relay or DOM).
| Live Stats & Categories | Appearance & Detection Mode |
|---|---|
 |
 |
FB Feed Diet is designed with a local-first privacy architecture:
- Zero Data Collection: No user data, browsing history, or feed content is collected, tracked, or transmitted. Zero telemetry, no backend servers, no external network requests.
- 100% Local Storage: All preferences and filtering statistics are stored strictly on your local device (
chrome.storage.local). - Minimal Permissions:
- Host permission (
*://*.facebook.com/*): Only used to execute filtering logic on Facebook pages. storage: Stores local preferences and statistics.scripting: Synchronizes settings dynamically with already-open Facebook tabs.
- Host permission (
- Open Source & Transparent: All code is open source and verifiable.
- Privacy Policy: For full terms, please refer to PRIVACY.md.
- Supports Firefox, Google Chrome, Microsoft Edge, and other Chromium-based browsers.
- Facebook frequently updates its DOM structure, post formatting, and internal data structures, which may cause classifiers to break over time.
- Provides two classification engines: Relay (direct inspection of Facebook's raw GraphQL data) and DOM (broader rendered page scanning).
- If you encounter unclassified posts or anomalies, feel free to report them on Issues with the feed probe log attached.
This project was inspired by:
Facebook updates frequently, and keeping the filters working requires ongoing maintenance time.
If the time and frustration it saves you is worth far more than a $2–$5 cup of coffee, consider buying me a coffee to support the long-term maintenance of this project!
This project is licensed under the MIT License.