What this guide is really about
Toxic replies don't have to cost you your open account. Threads ships with a setting called Hidden Words that filters junk out of your replies, and the automatic part is already on by default. You don't need to go private to get a calmer comment section.
The feature catches more than slurs. Custom filters accept words, hashtag topics, phrases, numbers, and emojis, so spoilers and seasonal noise are fair game too. Hidden content isn't deleted, it's covered, and you can lift the cover any time you want to see what got filtered.
Here's the full loop: confirm the automatic filter, build a custom list that fits your season, peek under the covers, and flip everything off when you're done. No private account required.
Hidden Words lives under Profile, Menu, Content preferences, Hidden Words. The automatic filter for common offensive words in replies is on by default and moves flagged replies under a cover at the bottom of your thread. Custom filters hide words, phrases, numbers, and emojis from your feeds, search, and profiles, with an optional duration. Hidden replies are not deleted: others can still view them.
What Hidden Words filters, and what it leaves alone
The documented app path, plus the web option for manual hides
How to see hidden replies and reverse any filter
The automatic filter for offensive, misleading, or spam replies is on by default; flagged replies move to a covered section at the bottom of your thread.
Custom filters accept words, hashtag topics, phrases, numbers, and emojis, and hide matching posts from your feeds, search results, and others' profiles.
The verified path is Profile, Menu, Content preferences, Hidden Words, all inside the Threads app.
Hidden is not deleted: others can still view covered replies, and messages you receive aren't filtered at all.
You can switch off the automatic reply filter, choose a duration for a new custom filter, or unhide a reply you hid manually. automatic filter off.
What Hidden Words Actually Filters
Hidden Words runs on two tracks. The automatic one is on by default: replies to your posts that contain words, phrases, or emojis which may be offensive, misleading, or spam get hidden without you lifting a finger. Flagged replies drop to a covered section at the bottom of your thread rather than disappearing.
The second track is yours to build. Custom filters take words, hashtag topics, phrases, numbers, and emojis, then hide posts that include them from your feeds, cover them in search results and on others' profiles, and tuck matching replies under a Hidden replies cover at the bottom of your post.
The Threads Help Center guide to Hidden Words spells out the part most people miss: hidden isn't deleted. Anyone can still open the cover and view a hidden reply, which keeps your thread readable for everyone else while sparing you the noise. The footprint has grown, too: in April 2024, Meta expanded Hidden Words beyond replies into feeds, search, and profiles, as CNET reported at the time.
The two tracks also differ in who they protect. The automatic filter applies to replies on your own posts, so it cleans the conversation under things you published. Custom filters follow you around instead: they shape what you see in your home feeds, in search results, and on other people's profiles, wherever the filtered items show up.
Turn It On: The Verified Path
Everything lives in the Threads app. Open your profile from the bottom right, tap the Menu at the top right, choose Content preferences, then Hidden Words. The first toggle, Hide common offensive words in replies, is on by default, so most of the job is confirming it stayed on.
Below the toggle sits Custom filter, home for your own lists. On the web, threads.com from a computer adds a manual cousin: open Options next to a reply and choose Hide for everyone. Others can still view hidden replies. That one's a separate mechanic from Hidden Words, built for single replies rather than patterns. The app path above follows Meta's Hidden Words help page, so it matches what you'll see in the app today.
One distinction worth locking in: Hidden Words filters what appears around your posts and in your feeds, not who can follow you. If your real goal is limiting the audience itself, make your Threads account private instead. That lever trades reach for quiet in a way word filters never will, and running both at once is perfectly normal.

Build a Custom Filter That Fits
Custom filters are where Hidden Words gets personal. From the Hidden Words screen, tap New filter below Custom filter. Enter a name and a description, tap Add words, and add whatever's polluting your replies: words, hashtag topics, phrases, numbers, or emojis. Under Choose how long to hide, tap Select to pick a duration, then tap Save.
The duration matters more than it looks. Say a show's season finale airs this week: a filter named after the show, loaded with character names and plot phrases, set to expire once the finale window passes, handles the spoiler rush and then retires itself. The same shape works for a politics-heavy news cycle or the engagement-bait phrases that pile on when a post takes off. That's an example, not a template: build lists from words you actually saw, not words you fear.
Filters got sharper in June 2025, when Threads picked up Hidden Words settings separate from Instagram's and began to let you set time limits on word filters, as TechCrunch reported. Before that split, your Instagram list did double duty across both apps. If an older walkthrough sent you hunting through Instagram settings, that's why.
Word filters pair naturally with other quieting settings. If you want a calmer-looking profile rather than calmer replies, you can hide likes on Threads too. Different screen, same instinct: fewer loud signals competing for attention.
Treating hidden as deleted. Covered replies stay viewable to everyone; filtering shapes what sits front and center, not what exists.
Expecting filters to cover messages. Current official wording says Hidden Words preferences won't apply to messages you receive on Threads.
Word-filtering a person problem. One account flooding every post calls for restrict or block, not a longer word list.
Following stale walkthroughs. Threads got its own Hidden Words settings separate from Instagram in June 2025, so older guides point to a different route.
Skipping the duration. Choose a duration that fits the topic when creating a spoiler or event filter.
See What Got Hidden and Reverse It
Hidden replies are covered rather than deleted. To see replies the automatic filter flagged, scroll to the bottom of the thread and open the covered section; hidden replies wait there, fully viewable by you and by anyone else who looks. Replies caught by custom filters sit under the Hidden replies cover on your post the same way.
Reversing is just as direct. A custom filter stops hiding content when you remove it or when its duration runs out. For a reply you hid manually with Hide for everyone, the official Threads account walkthrough says to find it in the Hidden replies list at the bottom of the thread, then open Options and choose Unhide. Meta's Hidden Words help page documents the covered replies and filter settings; the official Threads account walkthrough shows how to unhide a manually hidden reply.
A weekly audit keeps the lists honest. Glance under the covers on a busy thread: if real conversation is getting caught, trim the word list; if junk still slips through, add the exact phrases you're seeing. Topics drift, so treat filters as things you tune, not things you set once.

What Hidden Words Does Not Hide
The limits matter as much as the feature. Meta's current help page is blunt about the biggest one: "Your Hidden Words settings preferences currently won't apply to messages you receive on Threads." The automatic setting covers replies to your posts; custom filters also hide matching posts from your feeds and cover them in search results and on others' profiles. Your inbox is out of scope.
The covered-section model is the second blind spot. Hidden replies remain viewable to others, so filtering doesn't make harassing content invisible to the world; it moves it out of your line of sight. And because messages sit outside the feature, protecting your inbox means a different control entirely, starting with the option to turn off messages on Threads.
That points to the real decision: is the problem words, or is it a person? Spoilers, bait, and seasonal politics are word problems. One account flooding every post is a person problem, and when that's what you're facing, the choice is restrict vs block, not a longer filter list. Word-level for content, person-level for accounts, and messy threads often need both.
A quick test sorts it. Ask whether the annoyance follows a topic or an account. If the same spoiler phrase keeps arriving from different people, a filter solves it in one pass. If one account keeps showing up regardless of what you post, that account needs restrict or block. When both are true at once, run the filter for the topic and the person-level control for the account, then recheck the covered section in a week to see what each one actually caught.
A Five-Minute Settings Sweep
Here's the whole routine in one pass. Open Profile, Menu, Content preferences, Hidden Words. Confirm the offensive-words toggle is where you want it. Tap New filter, add the words you actually saw this week, select how long to hide them, and save. Done in the time a coffee takes to cool.
Afterward, open the covered section on a busy thread to check what was hidden. If you also want to put a number on how your posts are doing, use the Threads engagement rate calculator. That number measures engagement, not whether Hidden Words improved the conversation.
One more corner is worth sweeping while you're in there: the notification side. Filters cut what you see, not what pings you. If every reply alert is the real drain, you can turn off notifications on Threads separately. Five minutes, and the whole reply-noise stack is covered.
Action checklist
Use this as the practical next pass after reading the guide.
- +Open Profile, Menu, Content preferences, Hidden Words in the Threads app.
- +Confirm the Hide common offensive words in replies toggle is where you want it.
- +Create one custom filter from words you actually saw this week.
- +Set a duration under Choose how long to hide for anything time-boxed.
- +Check the covered section at the bottom of a busy thread to audit what got hidden.
- +After a week, review whether replies read better, then trim or extend your lists.

Conclusion
Hidden Words is the control to reach for first when reply noise climbs. It's on by default for the worst material, you can create custom filters with selected durations, and covered replies remain viewable. You can also switch off the automatic reply filter.
Set it once, tune it to whatever season you're in, and peek under the covers weekly. Your account stays open, your thread stays readable, and the junk stays out of sight without locking anything down.



