How to Use RSS Reader Keyword Filters to Cut the Noise
You subscribed to forty good feeds and now open your reader to 600 unread items, half of them about a product launch, a sports team or a sponsored roundup you will never read. Unsubscribing loses the good posts too.

You subscribed to forty good feeds and now open your reader to 600 unread items, half of them about a product launch, a sports team or a sponsored roundup you will never read. Unsubscribing loses the good posts too. RSS reader keyword filters solve this by hiding or marking items as read before they reach you, while keeping the source.
I have maintained the same reading setup for nine years, and filters are the reason it still works. This guide covers which readers support them, how to write rules that do not swallow the posts you want, and how to check whether a filter is helping or quietly hiding things.
What RSS reader keyword filters actually do
A filter is a rule that looks at each incoming item and takes an action when it matches. The match is usually on the title, sometimes on the body, author or category tags. The action is one of three things: mark as read, hide entirely, or highlight.
The difference matters. Marking as read keeps the item searchable later, so a false positive is recoverable. Hiding or deleting removes it from view, and in some readers it is gone for good. Start with mark as read every time.
Which readers support keyword filters
Support varies a lot, and several popular readers keep filters behind a paid tier. Here is where things stood in late 2026.
| Reader | Filter feature | Matches on | Cost |
|---|---|---|---|
| Miniflux | Block rules and keep rules per feed | Regular expressions on title, URL, content | Free if self-hosted, about $15 a year hosted |
| FreshRSS | Filter actions per feed or category | Search syntax such as intitle: and author: | Free, self-hosted |
| Inoreader | Rules and filters | Title, content, author, with many actions | Paid plans |
| Feedly | Mute filters | Keywords in title or content | Paid plans |
| NewsBlur | Intelligence trainer | Title words, authors, tags | Free tier, more on premium |
If you run your own server, Miniflux and FreshRSS give you the most control for no subscription fee. Our walkthrough on self hosting an RSS reader with FreshRSS or Miniflux covers the setup in an afternoon.
Writing filters that hide noise and keep the good posts
The trap with keyword filters is the short, common word. Filter "Apple" on a tech feed and you lose a great piece on supply chains that mentions it once. These are the rules I follow.
- Filter titles before bodies. A keyword in a title almost always means the post is about that subject. A keyword in the body often means nothing.
- Use phrases, not single words. "Sponsored post" or "deals of the week" are precise. "Deal" alone will catch a story about a merger.
- Use word boundaries in regex readers. In Miniflux, a block rule like (?i)\bpodcast episode\b matches the phrase in any case but not inside other words.
- Scope filters to the feed that causes the problem. A global rule for "review" will hurt the one blog where reviews are the reason you subscribed.
- Write down why each rule exists. I keep a plain text note with the date, feed and reason. Six months later you will not remember why "Week 34" is blocked.
Five filters worth setting up today
- Sponsored and affiliate posts: titles containing "sponsored", "partner content" or "deals".
- Recurring link dumps you already get elsewhere: "links for", "weekly roundup" from secondary feeds.
- Podcast episode notices on blogs that also publish articles: "episode" or "ep." at the start of the title.
- Event and webinar announcements: "webinar", "register now", "join us".
- One topic you are tired of for now, such as a sports season or a specific product line.
Why muting a whole topic is often the wrong move
A lot of advice says the cleanest reader is the one where you mute every topic you are not interested in. I disagree, and the reason is how curation actually works. The best items in my weekly link roundup often come from outside the subjects I thought I cared about. A broad topic filter on "AI" or "elections" removes the mediocre posts and the rare brilliant one together.
A better approach is to filter on format and source quality rather than subject. Block the press releases, sponsored posts and announcement stubs. Let substantive writing through on any topic. If a source keeps sending you subjects you do not want, it may simply be the wrong source, and vetting sources for a curated feed with a 30 day trial is a cleaner fix than a growing list of blocked words.
Here is a real example from my own reader. A design blog I follow publishes about 25 posts a week. Eight are sponsored tool roundups, five are job board reposts, and the rest are essays and case studies. Two title rules, one for "sponsored" and one for "now hiring", cut the feed to 12 items a week without touching a single essay. A topic rule on "tools" would have removed two of the best case studies I read that year.
The one exception is temporary muting. Blocking a news topic for two weeks during a story you cannot face is reasonable. Put an end date in your notes and remove the rule.
How to check that your filters are not hiding good posts
Filters fail silently. You never see what you missed, so you need a small audit habit.
Once a month, open the list of items your filters marked as read. Most readers let you view read items per feed or search for them. Skim titles for five minutes. If you find two or more items you would have opened, the rule is too broad. Narrow it to a phrase or limit it to one feed.
Also count. If one rule is catching more than 30 percent of a feed's items, ask whether you still want that feed at all. Keeping a source alive with heavy filtering often costs more attention than it saves.
When you find good posts that slipped through on the other side, star them. That habit feeds straight into the workflow in how I turn RSS stars into a weekly link roundup.
Frequently asked questions
Can you filter RSS feeds without a paid reader?
Yes. Miniflux and FreshRSS are free to self-host and both support filtering. NewsBlur offers basic training on its free tier. You can also run a feed through a self-hosted proxy that rewrites it, though that adds a moving part to maintain.
Do RSS filters work on old items already in my reader?
Usually not. Most readers apply rules only to items that arrive after the rule is saved. Some, including Inoreader, let you apply a rule to existing items manually, and FreshRSS can run filter actions against stored articles.
Should I filter by title or by full content?
Start with titles. Title matches are accurate and rarely catch posts you wanted. Content matching is useful only for specific phrases, such as a standard sponsorship disclosure line that appears in every paid post.
More in Curation
Curation
A Tagging System for Saved Articles That Survives a Year
I exported my read-later library last spring to check something unrelated and found 1,847 saved articles carrying 240 distinct...
Curation
How to Handle Paywalled Links in a Curated Reading List
The complaint that finally changed how I do this arrived in March, from a reader who had followed my weekly links for about two...
Curation
How to Share an OPML Bundle People Will Actually Import
Someone asked me in September for my feeds on soil science and small presses.