No AI LinkedIn Feed автор Philipp Karstaedt

No AI LinkedIn Feed helps you clean up your LinkedIn timeline by detecting posts that appear AI-written and hiding them based on your own comfort level.

Instead of a one-size-fits-all approach, the extension gives you full control over how detection works. We recommend a relatively cheap model such as the GPT-5 Nano, which already gives pretty good detection results.

Choose your detection provider:
- OpenAI API
- Anthropic API (Claude)
- Google Gemini API
- Local/self-hosted LLM endpoint

Then, configure the extension:
- Select a model for each provider (with custom model ID support for Anthropic and Gemini).
- Enter your own API keys directly in extension settings.
- Set a comfort threshold from 1–100 to decide how aggressively posts are filtered.
- Enable Test Mode to see scores and debug behavior before using full hiding mode.
- Toggle the extension on/off instantly without uninstalling it.

How it works:
When you browse LinkedIn, the extension evaluates post text and assigns an AI-likelihood score. If a post score is at or above your threshold, it is hidden (or labeled in Test Mode). Lower-scored posts remain visible.

Privacy and control:
Your provider keys and settings are stored in your browser’s extension storage. You decide which provider to use and where requests are sent (including local infrastructure, if preferred).