The flood of engagement bait —posts designed only to generate clicks, controversy, or shallow reactions— has become one of the biggest problems on X (formerly Twitter). In this context, Unbaited has emerged, an open-source browser extension that promises to bring calm back to your timeline using language models running on Groq’s ultra-fast API.
What is Unbaited?
Unbaited is a lightweight tool that integrates directly into your browser and analyzes tweets in real time as they appear in your feed. The system, powered by an LLM (Large Language Model) of your choice, can detect:
- Engagement bait posts (tweets asking for likes, retweets, or replies with tricks).
- Political or polarizing content that tends to inflame debates.
- Tweets designed to emotionally manipulate the reader.
When such content is detected, the extension automatically blurs the tweet. Users can still reveal it with one click, avoiding hard censorship while keeping the choice in their hands.
How it works
The process is straightforward:
- As you scroll through X, the extension detects new tweets in your viewport.
- It sends only the tweet’s text to Groq’s API for analysis.
- If flagged as bait or inflammatory, the extension blurs the tweet.
- With a single click, you can reveal it and decide whether to engage.
Important: currently, the analysis applies only to text. Images, videos, and links are not processed.
Installation
Chrome and Safari
- Install the extension from the Chrome Web Store.
- Get your Groq API key.
- Open the extension settings and enter your key.
- Optionally, customize the system prompt to fine-tune which content gets filtered.
Firefox
- Install the extension from Mozilla Add-ons.
- Enter your Groq API key in the settings.
- Optionally, adjust the system prompt for more tailored filtering.
Compatibility
- Chrome ✅
- Firefox ✅
- Safari ✅
All major modern browsers are fully supported.
For developers
The project is split into two parts:
extension/: the browser extension.landing/: the landing page built with Next.js.
Extension development
cd extension
npm i
npm build
Loading in browsers
Chrome/Safari:
- Open the extensions menu.
- Enable Developer Mode.
- Load the
extension/distfolder as an unpacked extension.
Firefox:
- Go to
about:debugging. - Select “This Firefox.”
- Load any file inside
extension/distas a temporary add-on.
Implementation notes
- Chrome/Safari: uses Service Workers (MV3).
- Firefox: falls back to Background Scripts.
The code handles these differences automatically, though developers should test functionality in each environment.
A step toward a healthier feed
At a time when algorithms prioritize interaction at any cost, Unbaited gives users some of that control back. It won’t eliminate all noise, but it provides a strong first line of defense against the toxic content that dominates feeds.
Frequently Asked Questions (FAQ)
Is Unbaited free to use?
The extension is open-source and free, but you’ll need a Groq API key. Depending on your plan, usage may incur costs.
Can I adjust what content gets filtered?
Yes, you can customize the system prompt to define which behaviors the AI should mark as bait or irrelevant.
Does it affect browser performance?
Thanks to Groq’s API speed, the impact is minimal, though it depends on your connection. For most users, browsing remains smooth.
What about my privacy?
Unbaited only sends tweet text to the API for analysis. It does not process personal data or access your X credentials.
