LESS AI FILLER. MORE OF YOUR INTERNET.

Keep the weird.
Skip the filler.

Blur likely AI-written posts on LinkedIn, X and Reddit. Keep the odd little stories, actual opinions and reasons you came here.

Add to Chrome, free Free forever.
No account needed for local checks.

Free on the Chrome Web Store · Works on LinkedIn, X and Reddit

A lime fishing hook carrying a paper strip that repeats the word engagementDon't take
the bait.
MADE FOR YOUR FEED
Available now

You came for
a break.
Not more noise.

A paper plane with a bright lime fold
A little room to wander.

One good post beats
twenty that say nothing.

The odd little story. A useful answer. An opinion that actually has a point. That's the internet you came for.

NPC Filter helps put likely AI-written posts out of your way. You decide what deserves a second look.

Try the filterNo install needed

You call
the scroll.

Blur it. Label it. Hide it.
Or read it anyway.

NPC Filter looks for signs of AI-written text. You choose how flagged posts appear. Every filtered post can be revealed.

Try the example here. This demo uses a preset result and sends no data.

YOUR CHOICETRY IT
Choose a filtering mode
JordanExample post

In today's fast-paced world, unlocking your potential is a game-changer.

Here are five ways to embrace the journey and take things to the next level.

It's not just a morning routine, it's a mindset shift, and it serves as a testament to the power of consistency.

Likely AI
Like   ·   Comment   ·   Share

A softer focus. The choice stays sharp.

Spot one we missed?
Tell us.

No filter catches everything. When a post or comment slips past, point at it. Reports help us tune the rules and spot patterns we miss on each site.

  1. See it. Hover a post or comment we did not flag and click the small Side-Eye icon.
  2. Tick what you noticed. Pick from five plain groups, with more specific signs under each. Add a short note if you like, up to 280 characters.
  3. It shapes the next update. We use reports to tune the rules and how much each one counts.

Flagged one by mistake? Press Wrong call? on it and tell us it got it wrong.

Reports never include the post's text or the author's name or handle. We get a fingerprint (hash) of the text, what you ticked, your note and NPC Filter's own scores. Keep the post itself out of your note. How we handle your data.

Free forever.
More if you want it.

Start with local filtering. Add deeper AI checks when you want another opinion on your feed.

THE EVERYDAY FILTER

Free. Actually free.

$0 / forever

Your five-minute break deserves something worth reading.

  • Unlimited local, rule-based checksNo account needed. Checked in your browser.
  • Rules that keep evolvingWe keep updating them as AI-writing patterns change.
  • Your choice: blur, label or hideReveal any post whenever you want.
  • 500 AI checks each monthIncluded when you sign in. No payment required.
Add to Chrome

No trial countdown. No card for Free.

NPC FILTER PRO

Small price. More peace.

$2.99 USD / month

Less than a fancy coffee.
A whole month of Pro.

  • Everything in FreeYour local filter keeps doing its thing.
  • 3,000 AI checks each monthSix times the signed-in Free allowance.
  • A second look beyond the rulesAI checks text the local rules haven't already flagged.
  • Blur likely AI-written contentOr choose label or hide. You're still in charge.
Add to Chrome

Then upgrade in the extension. Billed monthly.
Cancel any time from the extension.

Local checks stay free when your AI allowance runs out. AI checks run on our server; post text is sent without names or handles. How we handle your data.

A filter.
Not the final word.

How does the filter keep up?

The free filter checks writing patterns using local rules in your browser. We keep updating those rules as AI writing changes. It does not need an account or paid AI checks to work. No detector is perfect: you can always reveal a post or adjust the filter.

Is all AI writing bad?

No. AI-assisted writing can be useful, and human writing can be repetitive. Our starting point is giving you a choice over likely AI-written posts. We aren't deciding what everyone should value.

What if the filter gets it wrong?

It can. A "likely" flag is an estimate, not proof. You can reveal filtered content, change the mode, adjust sensitivity or pause the filter. Trust no one is the strictest setting and may flag more human writing.

Where does it work?

LinkedIn, X and Reddit, in Chrome. On LinkedIn it stays off until you turn it on. On X it is on by default and works on posts in timelines and on replies on a post's page. On Reddit it works on posts and their comments. Ads are always left alone.

How do I install it?

Add it from the Chrome Web Store. It is free to install and works in Chrome on LinkedIn, X and Reddit.

What happens to my data?

Most checks happen in your browser. If you sign in, posts the quick check does not already flag get a deeper check on our server. Only the words are sent, never names or handles. You can turn off anonymous usage stats. Read the privacy policy for the details.

What does it cost?

Local rule-based filtering is free forever, with ongoing rule updates and no account required. Sign in for 500 deeper checks a month, or go Pro for $2.99 a month and 3,000 checks. These are AI checks on our server. Pro is billed monthly in USD. Cancel any time from the extension. If your AI allowance runs out, local checks keep working.

How do I reach you?

Write to hello@npcfilter.com. We read every message. For privacy or account deletion, use privacy@npcfilter.com.

I like this place.
Just not
every post.

The niche obsessions. The terrible jokes. The posts only you would save. Keep what makes your feed yours.

Add to Chrome

Free forever local filtering. No account needed. Your feed, your call.

A folded paper heart with a lime-green fold
Keep what
you love.