What Is OPML? The File That Frees Your Feed List

The One File That Makes Feed Readers Non-Sticky — a painterly illustration conveying this idea from "The One File That Makes Feed Readers Non-Sticky"

Try to leave a social platform and take your feed with you. Not your list of followers — your actual feed, the ranked stream of posts you see when you open the app. There is no export button for that, because the thing that produces it isn’t a file you own. It’s a model running on someone else’s server, built from years of what you clicked, paused on, and scrolled past.

Now try to leave a feed reader. You open Settings, click “Export subscriptions,” get a file a few kilobytes in size, open a different reader, click “Import,” and you’re subscribed to the same sources again in under a minute. That file is called OPML, and the fact that it exists — plain, boring, and unremarkable — is the whole reason feed readers have never been able to lock anyone in the way social platforms do.

This isn’t a small technical footnote. It’s the concrete version of an argument this site makes about RSS generally: a subscription is something you hold, not something a platform grants you. OPML is where that claim stops being a slogan and becomes a file you can actually look at.

What’s Actually in the File

An OPML export isn’t a database dump or an account snapshot. It’s a text file — an outline, the same shape as a nested bullet list — where each line is either a folder or a feed. A feed line carries the feed’s web address and a human-readable title; a folder line just groups feed lines underneath it. That’s close to the entire contents. Open one in a text editor and you can read it in a minute, because there’s almost nothing to hide.

Worth being precise about what this article can and can’t tell you here: OPML’s own specification is published at opml.org, but the page renders its content client-side, which means its exact wording, version history, and authorship aren’t things this article can quote — we could not extract readable text from it. So rather than cite the spec directly, everything below about what OPML actually does is drawn from watching real exports move between real readers: FreshRSS, Miniflux, NetNewsWire, and NewsBlur all document OPML import and export in their own reference material, checked 2026-08-16. That’s a narrower claim than “here’s what the spec says,” but it’s the one this article can stand behind.

The important structural detail is that folders travel with the feeds. Because a folder is just an outline item that contains other outline items, your organization — “News,” “Tech,” “Newsletters,” whatever categories you built — survives the move along with the URLs themselves. You’re not re-typing forty feed addresses into a new account; you’re importing one file and your existing shelf reappears, sorted, in the new app.

The Five-Minute Reader Swap

This assumes you’ve already cleared two separate hurdles — finding a site’s feed in the first place, and picking which reader you want to use, both of which are their own decisions with their own trade-offs and get their own treatment elsewhere on this site. What follows is what happens once you’ve done both and decide to change your mind about the second one.

Here’s what actually happens when you switch, described concretely rather than abstractly:

In your current reader, find Settings → Import/Export (the wording varies slightly by app, but every mainstream reader puts it near account or subscription settings).
Export your subscriptions as OPML. You’ll get a single .xml or .opml file, typically under 100KB even for a few hundred feeds.
Open the reader you’re moving to, find its own import option, and point it at that file.
Your folders and feeds reappear. Depending on the reader, it may take a minute or two to fetch each feed’s current content for the first time.

The interesting part isn’t the mechanics — it’s that step 3 works at all regardless of which two readers are involved. As of the most recent check (2026-08-16), self-hosted options like FreshRSS and Miniflux and desktop/mobile apps like NetNewsWire all document OPML import and export as a standard feature, and hosted services Inoreader and NewsBlur do the same. NewsBlur’s own documentation goes further than most: it still accepts OPML exports that originated from Google Reader, a product that shut down in 2013. Thirteen years on, a file format outlived the company that generated it, the reader that replaced it, and probably the device it was first opened on.

Pricing and version numbers move fast in this space and this article won’t pretend otherwise — treat anything below as dated the day it was checked, not as a permanent fact. As of 2026-08-16, Miniflux’s hosted tier ran $15/year, NewsBlur’s paid tiers ranged from a free option up to $36–$99/year depending on archive depth, and Inoreader’s paid tier ran roughly $7.50–$10/month. None of that changes the argument here: what matters isn’t what any single reader charges, it’s that none of them can charge you for the exit.

Why This Keeps the Reader Market Honest

Compare the two situations directly. If a feed reader raises its price, ships an interface you dislike, or just quietly gets worse, your response is a five-minute file transfer and you lose nothing that matters — your sources come with you. If a social platform does the same thing, your response is starting over: rebuilding a following, re-training a recommendation model on a fresh account, accepting that the thing that made your experience good — the tuned, personalized ranking — doesn’t export and can’t be reconstructed from the outside. One market has switching costs near zero. The other has switching costs so high that people tolerate years of a product getting worse rather than pay them.

That asymmetry is why the reader market has stayed genuinely competitive — self-hosted, hosted, and local options coexisting, priced against each other, none of them able to assume you’re stuck. It’s not because feed reader companies are more virtuous than social platforms. It’s structural: the thing your experience is built from (a list of URLs) is small, portable, and yours by construction, instead of large, opaque, and generated server-side. A company that wants to keep you has to keep earning it, because the file that would let you leave is one settings menu away, in every reader that has ever mattered.

Two More Things to Do With the Same File

Migration is the obvious use, but the same small, boring file is useful twice more, and neither use gets much attention.

As a backup. An OPML export is a complete, human-readable record of every source you follow, independent of any account. Export one occasionally and keep it somewhere durable — it’s a hedge against a reader shutting down, a subscription lapsing, or an account getting locked, none of which are hypothetical in a market with hosted vendors that sometimes fold or pivot. Firefox’s own move away from built-in feed reading is a real precedent for this: when Mozilla retired feed preview and Live Bookmarks in Firefox 64 (11 December 2018), the documented escape hatch for existing users was exporting their bookmarked feeds to OPML for import into a third-party reader. The file was the continuity plan.

As a shareable reading list. Because OPML is just a file, it travels the way any file travels — email attachment, a link on a blog post, a gist. A journalist covering a beat, a researcher tracking a field, or a teacher building a course can hand someone a single .opml file of curated sources and have every recipient subscribed to the exact same list in seconds, folders and all. There’s no equivalent for “here’s my carefully tuned algorithmic feed, subscribe to my version of it” on any platform that ranks content for you — the closest thing is a public account someone can follow, which imports the person, not the reading list, and definitely not the editorial judgment behind it.

What OPML Doesn’t Save You

The honest limit matters as much as the portability claim, and skipping it would make this a worse article. OPML carries your subscriptions — nothing else. When you export and re-import, here is what does not come with you:

Read and unread state. Every article you’ve already read reappears as unread in the new reader. There is no OPML field for “seen.”
Stars, saves, and highlights. Anything you flagged for later in your old reader stays behind.
Tags and per-item labels. OPML preserves folder structure, but not the finer-grained tags some readers let you attach to individual stories.
Training and automation. NewsBlur’s per-story “intelligence” training and Inoreader’s automation rules are both real ranking layers a reader can build on top of a plain subscription list — and neither one is part of the export. You keep the sources; you lose the tuning.

So a reader swap genuinely costs you something: the accumulated read history and personal filtering you built up disappears, even though every source keeps flowing on day one in the new app. That’s a real trade, not a rounding error, and it’s worth weighing before you switch a reader you’ve used for years. What it isn’t is lock-in — you’re choosing to start your history over, not being prevented from leaving. Nobody has to re-earn your subscriptions from scratch, only your habits.

Frequently Asked Questions

Is OPML the same thing as RSS? No. RSS (and Atom) describes the content of a single feed — the articles themselves. OPML describes your list of feeds — which sources you subscribe to and how you’ve organized them. A reader uses OPML to know what to go fetch, then uses RSS or Atom to fetch it.

Does every feed reader support OPML? Every mainstream reader checked for this article — FreshRSS, Miniflux, NetNewsWire, Inoreader, and NewsBlur — documents both import and export as of 2026-08-16. It’s close to a de facto requirement for a reader to be taken seriously; a reader that locked subscriptions inside its own account with no export path would be a notable outlier.

Can I open or edit an OPML file myself? Yes. It’s plain text, structured the same way as a nested outline, and readable in any text editor. You can add a feed by hand, delete one, or merge two exports into one file before importing — no special software required.

Does exporting OPML back up the articles I’ve already read, or just the subscription list? Just the subscription list — the sources, not the content. If you want the articles themselves preserved, that’s a different, much larger backup problem, and it’s specific to whichever reader’s archive you’re using.

If OPML solves reader lock-in, why doesn’t every kind of app have something like it? Because OPML is only possible because a subscription is small — a URL and a folder label. An algorithmic feed’s real substance is a model trained on your behavior over months or years; there’s no compact file that represents it, and no incentive for the platform that built it to make leaving easy even if there were.