The mute and hide tools built into X, LinkedIn, and Reddit block words, accounts, and communities. None of them can tell how a post was written. To block AI-generated posts, you need a tool that reads each post and estimates whether it was likely written by AI. Botfilter is a free browser extension that does this on your device, with no account: it outlines, collapses, or hides likely AI-written posts and replies as the feed loads.
The Botfilter home page shows how flagged posts look on a demo feed. Platform steps below come from each site’s help pages, read on September 28, 2026.
What X, LinkedIn, and Reddit let you mute or hide
All three sites filter by something you name in advance.
X (formerly Twitter): muted words
X lets you mute posts that contain particular words, phrases, usernames, emojis, or hashtags. Muting a word also mutes its hashtag. On the web:
- Click More in the side menu, then Settings and privacy.
- Open Privacy and safety, then Mute and block, then Muted words.
- Click the plus icon and enter one word or phrase. X accepts one entry at a time.
- Choose Home timeline, Notifications, or both, and how long the mute lasts: forever, 24 hours, 7 days, or 30 days.
Muted posts drop out of your Home timeline, notifications, and replies. X’s help page says you will still see them in search.
LinkedIn: not interested, unfollow, and hide posts by a person
Click the More icon at the upper right of a post. Not interested removes that post, and LinkedIn says your feedback may be used to improve future feed recommendations. Where LinkedIn offers it, Hide posts by [name] or Unfollow reduces posts from one person, Page, or company. The author is not notified, and hiding a post does not report it.
Reddit: mute communities and block users
Reddit lets you mute up to 1,000 communities, from the community page, from a recommendation, or from the Muted communities list in settings. Muted communities stop appearing in your Home and Popular feeds and your notifications, though you can still visit them. You can also block up to 1,000 accounts. A blocked account’s posts stop showing up in your feeds, and its comments appear collapsed when you come across them.
| Site | Tool | What it filters on |
|---|---|---|
| X | Muted words | Words, phrases, usernames, emojis, and hashtags you enter |
| Not interested | One post, plus feedback that may shape future recommendations | |
| Unfollow or hide posts by a person | Everything from one person, Page, or company | |
| Mute a community | One community’s posts in Home, Popular, and notifications, up to 1,000 communities | |
| Block a user | Everything from one account, up to 1,000 accounts |
Sources: X, LinkedIn, and Reddit help pages, checked September 28, 2026.
Why mutes and blocks miss AI-written posts
A muted phrase catches only the posts that use that exact phrase, and it also hides every person who writes it themselves. AI-written posts rarely repeat the same wording, so a list of phrases stays one step behind the AI slop that fills replies. Blocking works one account at a time, and only after you have already read the post. Muting a community drops everything it posts from your feeds, including the people you wanted to hear from. To judge how a post was written, a filter has to read the post itself.
Other extensions that hide AI posts
A few browser extensions already try this. They differ most in how they decide and where your post text goes. The table uses each tool’s own site, read on September 28, 2026.
| Tool | How it decides | Where post text goes | Account or cost |
|---|---|---|---|
| Zenfeed | Not described on its site. It also filters keywords and promoted posts on LinkedIn and X. | Stays in your browser, per its site | Free, no account |
| NoSlop for LinkedIn | Eleven pattern rules, such as contrast hooks and emoji bullets | Stays in your browser, unless you turn on its optional Slop Score, which sends post text to TypeSafe’s Jev model | Free, no account |
| Slop Filter | Questions answered by a remote model, plus local punctuation counts | Sent to the provider you choose, TypeSafe or OpenRouter | Your own API key; its README puts 1,000 posts at about 4 cents |
| Pangram feed scanning | Pangram’s hosted detector | Sent to Pangram; feed data is stored by default for its Feed Health view, with an opt-out | Paid Individual plan, listed at $20 a month |
| Botfilter | A trained detection model, 23.6 MB, running in the browser | Stays in your browser | Free, no account |
Sources: each tool’s own site or README, checked September 28, 2026. Botfilter facts come from its current build. We have not tested the other tools, so the table compares how they work, not how accurate they are.
Of these, Botfilter is the only one we found, as of September 2026, whose site describes a trained model that runs in the browser and works beyond LinkedIn and X. If you are weighing a hosted detector for checking documents instead, see a free, private Pangram alternative for everyday reading.
Set up Botfilter in three steps
Botfilter scores text with a small detection model, a 23.6 MB file that ships inside the extension and runs in your browser. You don’t sign in or paste anything.
- Add Botfilter to Chrome from the Chrome Web Store. Botfilter is in development, and its listing is coming soon.
- Choose how flagged posts look. Click the toolbar button and pick a mode under Show AI text as: Outline, Collapse, or Hide. The mode applies on every site.
- Choose how many posts to flag. Under Flag, pick Fewer, Balanced, or More. Balanced is the default. A change applies to the open page right away, without checking the posts again.
If a page was open before you added Botfilter, reload it or click Scan this page in the popup. The popup also shows how many texts on the page it marked as likely AI-written, how many it checked, and how many are still in the queue.
Turn Botfilter off on one site
On the site you want to leave alone, click the toolbar button and clear On this site. Botfilter stops checking that site and removes its marks. To turn it back on, check the box again, or open Botfilter’s settings and click Turn on next to the site under Turned off on. The settings page also has a switch that turns Botfilter off everywhere.
Outline, Collapse, and Hide in your feed
Pick the mode by how much of a flagged post you still want to see. The three illustrations below show an X-style timeline at Balanced.
Spent all Saturday fixing the fence and the dog found the new gap in about four minutes. Back to the hardware store.
Most people will scroll past this. The ones who don't will be 10x ahead in 2027.
Here's the truth: success isn't about talent. It's about consistency, clarity, and relentless execution.
I've distilled 7 lessons from building 3 companies into one simple framework. 🧵👇
our landlord finally replaced the boiler, which means I no longer shower like a raccoon at 6am hoping for three minutes of hot water. small wins. ordering celebratory dumplings tonight
Unpopular opinion: AI won't replace you. But someone who uses AI will.
The key is to embrace change, stay curious, and never stop learning. The future belongs to those who adapt.
What's one skill you're investing in this year? Drop it below 👇
Wrapped our team offsite today. Biggest takeaway for me: we spend far too much time planning the work and not enough time finishing it. Trying two meeting-free afternoons a week and will report back.
Took the long way home along the river and the whole sky went this color for about ten minutes. Nobody else on the path even looked up, which felt like a secret.
Outline suits you if you want to read everything and decide for yourself. If a flag is wrong, click Not AI in the hover label and Botfilter stops marking that text.
Spent all Saturday fixing the fence and the dog found the new gap in about four minutes. Back to the hardware store.
Most people will scroll past this. The ones who don't will be 10x ahead in 2027.
Here's the truth: success isn't about talent. It's about consistency, clarity, and relentless execution.
I've distilled 7 lessons from building 3 companies into one simple framework. 🧵👇
our landlord finally replaced the boiler, which means I no longer shower like a raccoon at 6am hoping for three minutes of hot water. small wins. ordering celebratory dumplings tonight
Unpopular opinion: AI won't replace you. But someone who uses AI will.
The key is to embrace change, stay curious, and never stop learning. The future belongs to those who adapt.
What's one skill you're investing in this year? Drop it below 👇
Wrapped our team offsite today. Biggest takeaway for me: we spend far too much time planning the work and not enough time finishing it. Trying two meeting-free afternoons a week and will report back.
Took the long way home along the river and the whole sky went this color for about ten minutes. Nobody else on the path even looked up, which felt like a secret.
Collapse keeps your place in the feed. Each bar carries the post’s own probability, so you can still open anything that catches your eye.
Spent all Saturday fixing the fence and the dog found the new gap in about four minutes. Back to the hardware store.
Most people will scroll past this. The ones who don't will be 10x ahead in 2027.
Here's the truth: success isn't about talent. It's about consistency, clarity, and relentless execution.
I've distilled 7 lessons from building 3 companies into one simple framework. 🧵👇
our landlord finally replaced the boiler, which means I no longer shower like a raccoon at 6am hoping for three minutes of hot water. small wins. ordering celebratory dumplings tonight
Unpopular opinion: AI won't replace you. But someone who uses AI will.
The key is to embrace change, stay curious, and never stop learning. The future belongs to those who adapt.
What's one skill you're investing in this year? Drop it below 👇
Wrapped our team offsite today. Biggest takeaway for me: we spend far too much time planning the work and not enough time finishing it. Trying two meeting-free afternoons a week and will report back.
Took the long way home along the river and the whole sky went this color for about ten minutes. Nobody else on the path even looked up, which felt like a secret.
Hide is the strictest. Flagged posts leave the page, and switching back to Outline or Collapse brings them back.
On X
A reply is its own post, so each reply in a thread is scored and marked on its own. Botfilter skips any text under 25 words, on every site. On X, many posts and replies are shorter than that, so they are left unmarked.
On LinkedIn
Here, Botfilter marks the whole feed card, and it scores each comment separately from the post above it. A flagged comment under a human post is folded or hidden without touching the post.
I got rejected from 47 jobs in 2019.
Today I run a team of 60.
Here's what nobody tells you about rejection:
It isn't a stop sign. It's a redirection.
Every no taught me something: → Resilience is a muscle, not a trait. → Your network is your net worth. → Growth lives outside your comfort zone.
The truth is, the most successful people I know didn't avoid failure. They embraced it, learned from it, and kept going.
So if you're facing rejection right now, remember this: your breakthrough might be one application away.
Agree? ♻️ Repost to help someone in your network who needs to hear this today.
We're hiring a second-shift baker. Four nights a week, 10pm to 6am, and yes, you get to take home whatever doesn't sell. You don't need a résumé. Come by Thursday afternoon and bake something with us. Tell anyone who might like the quiet hours.
Excited to share that our team just wrapped up an incredible quarter. 🎉
This journey has been a true testament to the power of collaboration, innovation, and a shared commitment to delivering value for our customers.
A few key takeaways: • Alignment across teams is everything • Customer feedback is a gift • Small wins compound over time
I'm incredibly grateful to work alongside such talented, passionate people who show up every day ready to make an impact. Here's to continuing to push boundaries and redefine what's possible in the months ahead.
Last week our warehouse ran its first full inventory count without overtime. It took three tries to get the process right. The first attempt fell apart because nobody owned the spreadsheet. The second time we split the building into zones and gave each shift lead one zone and one scanner. By the third count people were finishing early and asking what else needed doing. Nothing fancy, only clear ownership and fewer handoffs. Sharing in case another ops team is stuck in the same loop.
I got rejected from 47 jobs in 2019.
Today I run a team of 60.
Here's what nobody tells you about rejection:
It isn't a stop sign. It's a redirection.
Every no taught me something: → Resilience is a muscle, not a trait. → Your network is your net worth. → Growth lives outside your comfort zone.
The truth is, the most successful people I know didn't avoid failure. They embraced it, learned from it, and kept going.
So if you're facing rejection right now, remember this: your breakthrough might be one application away.
Agree? ♻️ Repost to help someone in your network who needs to hear this today.
Congrats on the team, Daniel. The 2019 part hit home for me.
We're hiring a second-shift baker. Four nights a week, 10pm to 6am, and yes, you get to take home whatever doesn't sell. You don't need a résumé. Come by Thursday afternoon and bake something with us. Tell anyone who might like the quiet hours.
Excited to share that our team just wrapped up an incredible quarter. 🎉
This journey has been a true testament to the power of collaboration, innovation, and a shared commitment to delivering value for our customers.
A few key takeaways: • Alignment across teams is everything • Customer feedback is a gift • Small wins compound over time
I'm incredibly grateful to work alongside such talented, passionate people who show up every day ready to make an impact. Here's to continuing to push boundaries and redefine what's possible in the months ahead.
Last week our warehouse ran its first full inventory count without overtime. It took three tries to get the process right. The first attempt fell apart because nobody owned the spreadsheet. The second time we split the building into zones and gave each shift lead one zone and one scanner. By the third count people were finishing early and asking what else needed doing. Nothing fancy, only clear ownership and fewer handoffs. Sharing in case another ops team is stuck in the same loop.
On Reddit
Botfilter scores a post’s title and body together and marks the whole post in the feed. Each comment body is scored on its own, so a flagged comment can fold under a human post. Old Reddit works the same way, except that it scores the post body without the title. The illustrations on this page show X-style and LinkedIn-style feeds; the three modes work the same way on Reddit.
Switch between both illustrated feeds, each mode, and each sensitivity here:
Balanced flags about 1 in 100 human posts by mistake.
Spent all Saturday fixing the fence and the dog found the new gap in about four minutes. Back to the hardware store.
Most people will scroll past this. The ones who don't will be 10x ahead in 2027.
Here's the truth: success isn't about talent. It's about consistency, clarity, and relentless execution.
I've distilled 7 lessons from building 3 companies into one simple framework. 🧵👇
our landlord finally replaced the boiler, which means I no longer shower like a raccoon at 6am hoping for three minutes of hot water. small wins. ordering celebratory dumplings tonight
Unpopular opinion: AI won't replace you. But someone who uses AI will.
The key is to embrace change, stay curious, and never stop learning. The future belongs to those who adapt.
What's one skill you're investing in this year? Drop it below 👇
Wrapped our team offsite today. Biggest takeaway for me: we spend far too much time planning the work and not enough time finishing it. Trying two meeting-free afternoons a week and will report back.
Took the long way home along the river and the whole sky went this color for about ten minutes. Nobody else on the path even looked up, which felt like a secret.
What Botfilter checks on each site
Botfilter has a layout for each supported site, so it marks the unit you read: a whole post, a comment, or an email. It scores only the text. Images and video are not checked, and text you are typing is never read.
| Site | Scored | Marked |
|---|---|---|
| X | The post text | The whole post, including replies, each on its own |
| Post text and comment text | The whole feed card; each comment separately | |
| Post title and body; comment body | The whole post; each comment body | |
| Old Reddit | Post and comment bodies | Each post and comment body |
| YouTube | Comments and the video description | Each comment; the description |
| Hacker News | Comments and Ask HN or Show HN text | Each comment; the post text |
| Gmail | The open message; inbox subject and preview | The message body; inbox rows get a stripe, dimmed in Collapse and Hide |
| Any other site | Paragraphs, list items, and quotes | Each one on its own |
Source: the site layouts in Botfilter’s current build.
Email works a little differently, since folding inbox rows would break the list. The details are in how to spot and filter AI-written email in Gmail.
Choose how many posts to flag
Sensitivity trades wrong flags on human posts against missed AI posts. Botfilter set each level to flag a target share of human writing by mistake: about 1 in 200 at Fewer, 1 in 100 at Balanced, and 3 in 100 at More. The table shows how that played out on its sealed test split, a set of texts held out from training and scored once.
| Text | Texts | Fewer | Balanced | More |
|---|---|---|---|---|
| Short AI-written social posts, flagged | 658 | 94.2% | 95.9% | 98.3% |
| Human Reddit posts written before ChatGPT (2006 to 2016), flagged by mistake | 8,000 | 0.2% | 0.5% | 1.2% |
| Human web pages written before 2020, flagged by mistake | 8,000 | 0.9% | 1.8% | 5.1% |
Source: Botfilter v5 test results, published on the model card. Short social posts are AI-written posts of 25 to 80 words.
The web pages row matters on sites without their own layout, where Botfilter checks text paragraph by paragraph. On AI-written text from Gemini 3.8 Flash, a model it was never trained on, it flagged 94% of 147 texts at Balanced. How we trained Botfilter has every source and rate.
Start at Balanced. Move to Fewer if Botfilter folds posts you know a person wrote. Move to More if too many AI-written replies still get through, and pair it with Outline or Collapse so a wrong flag costs you a click instead of a post.
Every row above comes from sources and AI models Botfilter was trained on, with the test texts kept out of training, so rates on your own feed may be worse. Human posts from X and LinkedIn were not available to test, so how often it flags real people there is unmeasured. It was trained and tested on English text. Human writing can be flagged, and a post a person wrote with AI help is often missed. Each mark is a probability, not proof of who wrote the post.
For what these rates mean in practice, see what published studies say about AI detector accuracy.
Post text stays in your browser, with no sign-in or API key
A feed filter reads every post you scroll past. Botfilter scores each one inside your browser, you don’t need an account, and page text never leaves your device. It never connects to your X, LinkedIn, or Reddit account and asks for no password or API key. It reads only the page you already have open, and it never posts, reports, or tells the author anything.
Unlike the feed tools above that send post text to a provider, Botfilter makes no network requests for detection and has no server. How Botfilter runs AI detection on your device covers what it stores, including synced settings, and how to check its network use. The privacy page lists what is stored.
Use mutes and Botfilter together
Keep the site’s own mutes and blocks for topics, people, and communities you never want to see, and let Botfilter handle likely AI-written posts from everyone else. For the story behind it, read Introducing Botfilter.
Congrats on the team, Daniel. The 2019 part hit home for me.