Find the RSS feed for any website

How to find a site's RSS feed

1. Paste any page from the site: the homepage, a blog section, a YouTube channel or video, a subreddit, a Medium profile or a GitHub repository. You can leave out https://.
2. Press Find feeds. The tool collects candidate feeds and fetches each one to check that it really is a feed.
3. Copy the feed URL into your reader. Each result shows the feed's format, how many items it carries and when the newest one was published, so you can tell a live feed from an abandoned one before you subscribe.

Most publishing platforms emit a feed even when nothing on the page mentions it. WordPress, Ghost, Substack, Blogger, Medium and most static-site generators all do. That is why a site with no visible RSS icon usually still has a working feed one path away.

How the feed finder works

1. Platform patterns. For sites whose feed addresses follow a fixed pattern, the tool builds the feed URL directly: YouTube channels and playlists, subreddits and Reddit users, Medium profiles, publications and tags, Substack newsletters, Blogger blogs, and GitHub release and tag feeds.
2. Declared feeds. It reads the page's HTML and collects every <link rel="alternate"> tag whose type is RSS, Atom or JSON Feed, the standard way for a page to announce its feeds to browsers and readers. A blog often declares several: posts, comments, a podcast, a category.
3. Linked feeds and common paths. Only if nothing declared works does it look for feed links in the page body (an RSS icon, a "Subscribe" link) and then probe the conventional locations: /feed, /rss, /rss.xml, /feed.xml, /atom.xml, /index.xml and /feed.json, first under the section you pasted and then at the site root.
4. Validation. Every candidate is downloaded and parsed. A URL only counts as a working feed if it actually contains RSS, Atom or JSON Feed, not an error page or a homepage served at the feed's address, which is a common way for "feed" URLs to fail silently.

Feed URLs for popular platforms

If you prefer to build a feed URL by hand, these are the patterns the tool uses. Each was checked against a live site.

• WordPress: add /feed/ to the site address. Comments are at /comments/feed/, and categories, tags and authors have their own feeds, such as /category/news/feed/, /tag/themes/feed/ and /author/name/feed/, as documented in WordPress.com's feed guide.
• YouTube: https://www.youtube.com/feeds/videos.xml?channel_id=UC… for a channel and …?playlist_id=PL… for a playlist. The channel ID is the 24-character code starting with UC; YouTube's own channel pages declare this feed address. If you only have an @handle, the YouTube channel ID finder turns it into the ID.
• Reddit: add .rss to a subreddit or user page: https://www.reddit.com/r/name/.rss or https://www.reddit.com/user/name/.rss.
• Medium: medium.com/feed/@username for a writer, medium.com/feed/publication-name for a publication, medium.com/feed/tag/tag-name for a topic, and medium.com/feed/publication-name/tagged/tag-name for one tag inside a publication, per Medium's help center, which also notes that stories behind the paywall are not available as full stories in feeds.
• Substack: add /feed to the newsletter's address, whether it is name.substack.com or a custom domain.
• Ghost: /rss/ on the site.
• Blogger: /feeds/posts/default on the blog's address.
• GitHub: github.com/owner/repo/releases.atom for releases, /tags.atom for tags and /commits/branch.atom for commits on a branch.
• Mastodon: add .rss to a profile, such as mastodon.social/@name.rss. Bluesky profile pages also declare an RSS feed of the account's posts, which the tool picks up.

RSS, Atom and JSON Feed: what the format label means

The result shows which of the three feed formats a URL serves. RSS 2.0 is the most common and is what WordPress, Substack and most podcasts publish. Atom is a stricter standard used by YouTube, GitHub, Reddit and many static-site generators. RSS 1.0 (RDF) still appears on some older sites, and JSON Feed is a newer format that expresses the same thing in JSON instead of XML. Every mainstream feed reader handles RSS and Atom; most handle JSON Feed too. When a site offers more than one format of the same feed, they carry the same posts, so pick either.

Reading the results

• Valid feed: the URL returned a real feed. The item count shows how many entries it currently carries (many sites keep only their most recent entries), and the newest date tells you whether the source is still publishing.
• Not verified: the address follows the platform's standard pattern, but the site refused or timed out our check, which bot protection often does to automated requests. It will usually still work in your reader.
• Not a working feed: the page declares this feed, but the address returns an error or something that is not a feed. The site's feed is broken or was moved; tell the site owner, or use another result.
• Where it was found: declared by the page, built from the platform's pattern, linked in the page body, or found at a common path. Declared and platform feeds are the most dependable; a feed found only by probing works today but is not one the site advertises.

Why a feed can't be found

• The site has no feed. Some platforms never offered one, and some sites turn feeds off. Nothing declared, nothing at the usual paths: that is a real answer.
• The site blocks automated requests. Sites behind aggressive bot protection may refuse our server even though their feed works in a reader. Open the page in your browser, view its source and search for application/rss+xml to find the declaration yourself.
• The page is built entirely by JavaScript. Single-page apps sometimes add their feed link only after scripts run. Try the site's blog or news section, which is more often a plain page.
• The content is behind a login. Feeds for private or paid content are usually personal URLs issued inside your account settings, and no finder can discover them.

After the feed: the part readers don't solve

A feed reader is good at telling you what is new and bad at telling you what you read six months ago. That is a different job: saving the handful of pieces worth keeping and being able to find them later. Marqly saves articles and videos into one library from the browser extension or the iOS and Android apps, and on Pro itssemantic search finds a save from a description rather than its title. If your reading queue is the problem, start with theread-it-later workflow; if you curate sources for a team or a newsletter, see content curation with Marqly. For link rot in an old list of sources, the dead link checker is next door.