Every feed format, one JSON shape
The RSS & Feed API is a universal RSS, Atom and JSON Feed reader that returns clean JSON.
5 active endpoints. Every call is 1 credit.
- POST/feed-gateway/v1/fetch
- POST/feed-gateway/v1/discover
- POST/feed-gateway/v1/merge
- POST/feed-gateway/v1/search
- POST/feed-gateway/v1/batch
What RSS & Feed endpoints does ReefAPI ship?
5 live read endpoints. Read-only data API: no writes, no account actions, no dashboard access on the target site.
RSS & Feed API
3 of 5 endpoints, ready to run
One feed normalised: the channel with its title, language and image, and items with id, title, link, published time, author and summary.
{ "ok": true, "meta": { "api": "feed-gateway", "endpoint": "fetch", "mode": "live", "latency_ms": 254.9, "record_count": 20, "cache_hit": false }, "data": { "feed": { "url": "https://feeds.bbci.co.uk/news/rss.xml", "final_url": "https://feeds.bbci.co.uk/news/rss.xml", "format": "rss2.0", "title": "BBC News", "link": "https://www.bbc.co.uk/news", "description": "BBC News - News Front Page", "language": "en-gb", "updated": "2026-08-29T14:49:11Z", "image": "https://news.bbcimg.co.uk/nol/shared/img/bbc_news_120x60.gif", "generator": "RSS for Node", "categories": [] }, "items": [ { "id": "https://www.bbc.co.uk/news/articles/c158939pz4do#0", "title": "Christian festival 'devastated' as man dies after sign falls", "link": "https://www.bbc.co.uk/news/articles/c158939pz4do?at_medium=RSS&at_campaign=rss", "published": "2026-08-29T14:12:30Z", "updated": null, "summary": "Police say at least six other people were also injured in the incident at the Big Church Festival.", "categories": [], "enclosures": [], "image": "https://ichef.bbci.co.uk/ace/standard/240/cpsprodpb/9ec8/live/5cdf5720-a383-11f1-b109-879e35c24276.png" }, { "id": "https://www.bbc.co.uk/news/articles/c86xwqez4npo#0", "title": "At least 37 dead and hundreds evacuated after strike on Kyiv weapons depot", "link": "https://www.bbc.co.uk/news/articles/c86xwqez4npo?at_medium=RSS&at_campaign=rss", "published": "2026-08-29T14:43:04Z", "updated": null, "summary": "An investigation has begun into the storage of weapons for Ukraine's military near a residential area.", "categories": [], "enclosures": [], "image": "https://ichef.bbci.co.uk/ace/standard/240/cpsprodpb/9694/live/9464f3f0-a3b8-11f1-a291-b542ee92de7c.jpg" }, { "id": "https://www.bbc.co.uk/news/articles/cvgy48r1nr4o#0", "title": "Burnham's summer tour is over - now he has to prove himself in Parliament", "link": "https://www.bbc.co.uk/news/articles/cvgy48r1nr4o?at_medium=RSS&at_campaign=rss", "published": "2026-08-29T10:31:18Z", "updated": null, "summary": "The imminent return of Parliament will put the new prime minister under real scrutiny, with MPs finally getting a chance to grill him.", "categories": [], "enclosures": [], "image": "https://ichef.bbci.co.uk/ace/standard/240/cpsprodpb/84e7/live/9e49e220-a2f7-11f1-8d5b-155abe464768.jpg" } ], "item_count": 20, "total_items": 29 } }
How the RSS & Feed API works
RSS & Feed is a normal ReefAPI surface — the same four rules that hold for every other engine on the key.
No OAuth app, no request signing, no per-site account. One key covers all 184 engines.
Every route is a POST with a JSON body. Parameters are validated against the published schema before anything is charged.
Credits, not seats. Failed and blocked calls are never charged, and cache hits cost nothing.
One envelope everywhere. meta carries latency_ms, record_count and the endpoint that answered.
One stream out of thirty sources
Feeds are the last genuinely open publishing format, and every one of them is a slightly different dialect of two and a half standards with dates in four formats.
{"url": "https://example.com", "probe": true}Finds what a site publishes, and probing confirms each candidate actually returns a feed rather than a 404 page.
{"urls": ["…", "…"], "dedup": true, "since": "2026-08-01"}Then everything in one time-sorted, deduplicated stream — one call instead of one per feed plus your own merge.
Dates arrive normalised to a single format regardless of what the feed used, which removes the most tedious part of consuming feeds at any scale.
curl -X POST https://api.reefapi.com/feed-gateway/v1/fetch \
-H "x-api-key: $REEF_KEY" \
-H "content-type: application/json" \
-d '{"url":"https://feeds.bbci.co.uk/news/world/rss.xml","limit":10}'{
"ok": true,
"data": { … },
"meta": {
"api": "feed-gateway",
"endpoint": "fetch",
"mode": "live",
"latency_ms": …,
"record_count": …
},
"error": null
}One item schema across three feed formats, and where each action's limits sit
The point of this engine is that an RSS 2.0 item, an Atom entry and a JSON Feed object come back with the same keys, so you write one parser. What is not normalized is the item id, which stays in the source feed's own shape. This table records format detection, date handling, the dedup arithmetic and the per-action caps, all from live calls on 2026-08-27.
| Thing | Behavior | Measured |
|---|---|---|
| feed.format | Auto-detected, and also echoed as meta.format | The BBC world feed came back "rss2.0", github.com/python/cpython/releases.atom came back "atom", and daringfireball.net/feeds/json came back "json-feed". |
| items[].published | Always ISO 8601 UTC with a Z, whatever the source format used | "2026-08-26T11:32:11Z" from an RFC-822 RSS date, "2026-08-12T13:57:54Z" from an Atom RFC-3339 date, "2026-08-26T20:18:04Z" from JSON Feed. updated is frequently null even when published is set. |
| items[].id | The feed's own identifier, passed through, and the one field whose shape varies by source | RSS: the article link with a "#0" suffix. Atom: "tag:github.com,2008:Repository/81598961/v3.12.14". JSON Feed: the permalink URL. A Megaphone podcast: a bare UUID. |
| summary against content_html | summary is always clean plain text; the raw HTML body only arrives with include_content=true | content_html is capped at 30 KB. Leaving include_content off is the difference between a 20 KB response and a multi-megabyte one. |
| Podcast fields | Present only on iTunes-flavored feeds | The Vergecast returned feed.podcast {explicit: false, type: "episodic", owner: "Vox Media Podcast Network"}, item.podcast {episode_type: "full", duration_seconds: 1065} and enclosures[0].type "audio/mpeg". |
| merge dedup | Canonical link key with tracking parameters stripped; limit is applied after dedup | The same BBC feed under two URLs (rss.xml and rss.xml?v=2) gave total_before_dedup 54, unique_after_dedup 27, removed_duplicates 27. With dedup:false the same call gave 54 / 54 / 0. |
| since / until | ISO date or datetime, filtered before limit is applied | since 2026-08-26 on a 27-item BBC feed returned item_count 17 while total_items still reported 27. |
| Caps | fetch and search 100 items, merge 200, batch 20 feeds, merge and search 10 URLs, discover 20 feeds | Out-of-range integers are clamped rather than rejected: limit 500 returned ok:true with all 27 items the feed had. |
| Bad enum value | Rejected, with the legal values in the error detail | sort_by "bogus" returned INVALID_PARAM, message "invalid value 'bogus' for 'sort_by'", detail.allowed ["feed_order", "published", "title", "author"], retryable false. |
sort_by defaults to feed_order, so fetch, search and batch return items in the publisher's own order rather than newest-first. Pass sort_by=published when you need chronology. merge is the exception: it is date-sorted by construction and only takes sort_dir.
What normalising actually gets you
Measured across live feeds.
The three formats disagree about almost every field name, and publishers extend them freely. Items come back with the same keys regardless — id, title, link, published, updated, author, summary — so a reader needs one parser rather than three plus exceptions.
Feeds carry dates in several formats and frequently with inconsistent timezones. Everything comes back in one format, which is what makes sorting and since-filtering possible at all.
Passing many URLs returns one time-sorted stream with cross-feed duplicates removed, which is what happens when several feeds carry the same syndicated item. One call for a reader's whole subscription list rather than one per feed.
Candidate feed URLs are found from the site's markup and the conventional locations, and probing fetches each one to confirm it is live. A link tag pointing at a dead feed is common enough that the confirmation is worth the calls.
Many publishers deliberately publish a truncated summary to drive clicks. Content is returned when the feed carries it and requested, and where a feed only carries a summary that is what you get — pair it with an article extractor when you need the body.
What people build with RSS & Feed
The jobs this data is most often used for.
endpoints
credit per call
Feed readers call fetch to normalize RSS, Atom and JSON feeds into one schema.
Content pipelines use merge to combine many sources into a single sorted stream.
Monitoring tools use discover and batch to find and poll every feed for a set of sites.
What RSS & Feed data costs
The cheapest call here is 1 credit, so $15/mo (Pro) buys 10,000 of them — $1.50 per 1,000 credits. Credits roll over and never expire, and failed or blocked calls are not charged.
Full pricing →- 1,000 free credits on signup, no card
- One key, all 184 APIs, one credit pool
- Failed and blocked calls are never charged
- Credits roll over and never expire
Call it in two lines
Sign up, get 1,000 credits and one key that works on every engine. Then this is the whole protocol.
curl -X POST https://api.reefapi.com/feed-gateway/v1/fetch \
-H "x-api-key: $REEF_KEY" \
-H "content-type: application/json" \
-d '{"url":"https://feeds.bbci.co.uk/news/world/rss.xml","limit":10}'import requests
r = requests.post(
"https://api.reefapi.com/feed-gateway/v1/fetch",
headers={"x-api-key": REEF_KEY},
json={
"url": "https://feeds.bbci.co.uk/news/world/rss.xml",
"limit": 10
},
)
print(r.json()["data"])Have a question? We got answers.
The questions people actually ask before wiring up RSS & Feed.
Get a free key →I passed a website URL to fetch and got PARSE_ERROR.▾
fetch takes a feed URL, not a site URL. A fetch of https://example.com returned ok:false with error code PARSE_ERROR, retryable false, and the message "url returned an HTML page, not a feed, try the discover action". That is exactly the right next step: discover takes the site URL and hands back feed URLs you can then fetch.
Can discover find a feed on a site that blocks scrapers?▾
Often, yes, because it does not depend on reading the page. A discover call on reddit.com/r/programming reported page_status 403, meaning the HTML page itself was refused, and still returned https://www.reddit.com/r/programming.rss with 25 items after checking 18 candidates. The discovered_via field tells you which route worked: "link-tag" when the page declared the feed, which is how theverge.com resolved, and "probe" when a common path was guessed and then verified by parsing it.
What does discover return for a site with no feed at all?▾
Success with nothing in it. example.com returned ok:true, feeds [], count 0, candidates_checked 15 and meta.record_count 0. There is no error to catch, so branch on count rather than on ok. candidates_checked tells you how much was tried before giving up, which was 15 to 18 probes across the calls we made.
How does merge decide that two items are the same story?▾
On a canonical link key with tracking parameters stripped, falling back to guid or title. Measured directly: the same BBC feed requested under two different URLs produced total_before_dedup 54, unique_after_dedup 27 and removed_duplicates 27. The two genuinely different BBC feeds, world and technology, overlapped on nothing that day, giving 47 in, 47 out and 0 removed. Those four counters are in every merge response, so you can always see what was collapsed.
Both my feeds are called BBC News. How do I tell items apart?▾
Every merged item carries source_feed {url, title}, and the url is what distinguishes them. In a merge of the BBC world and technology feeds, both entries in the feeds[] array reported title "BBC News" while source_feed.url on each item pointed at the specific feed it came from. Key on source_feed.url, never on the title.
How does keyword search across feeds behave?▾
search accepts either url or urls (up to 10) and reports what it looked at. "trump ukraine" with match=any over the BBC world feed returned item_count 6 with total_scanned 27. "ai data" with match=all over BBC world plus the Hacker News frontpage returned item_count 1 with total_scanned 47. Matching is case-insensitive substring over title and summary by default; adding "content" to fields only helps if you also set include_content true, because otherwise there is no body to search.
Can I use search as a pure blocklist?▾
Yes. Pass an empty query with exclude set and it becomes an exclusion filter. Measured: an empty query with exclude="trump" over a 27-item BBC feed returned item_count 23 out of total_scanned 27, and echoed query back as null. That is the equivalent of an RSS filter-out rule, without an account.
What happens to one dead feed inside a batch?▾
It is isolated and reported, and the call still succeeds. A batch of three URLs where one host does not resolve returned ok:true with feed_count 3, ok_count 2 and meta.feeds_ok 2. The bad entry came back as {url, ok: false, feed: null, item_count: 0, error: {code: "NOT_FOUND", message: "host ... does not resolve"}}. Check each result's own ok flag rather than the outer one.
What is the RSS & Feed API?▾
RSS & Feed API is a ReefAPI endpoint group for rss & feed It returns live JSON through POST requests under /feed-gateway/v1.
Is the RSS & Feed API free to try?▾
Yes. ReefAPI starts with 1,000 free credits, no card required. RSS & Feed calls use the same shared credit balance as every other ReefAPI engine.
Do I need a RSS & Feed login or account?▾
No login to RSS & Feed is needed for the API response. You call ReefAPI with your x-api-key header, and the playground can run live examples before you create a production key.
How fresh is the RSS & Feed data?▾
The page example is captured from a live fetch call, and production requests fetch live data through ReefAPI rather than a static sample.
How many credits does the RSS & Feed API use?▾
RSS & Feed actions currently cost 1 credit per successful call. Failed or blocked calls are free, and all APIs draw from one credit pool.
Can I call RSS & Feed from an AI assistant or MCP client?▾
Yes. Connect ReefAPI once through MCP and your assistant can call feed-gateway actions with the same key, credit pool and JSON envelope used by normal REST requests.
25 Media, Film & Knowledge APIs on the same key
One key, one credit pool, one response envelope. If you are pulling RSS & Feed, you are one call away from the rest of the category — no second contract, no second integration.
Need something this API does not do?
Name the endpoint, the field, or a source we do not carry yet. We ship new APIs every week and you would be first to get the key. Real people read every message and reply the same day.
Try it on your own data before you pay anything
The call above is the real endpoint, not a recording. A free key gives you 1,000 credits, the other 183 APIs, and the same envelope everywhere.
Endpoints, parameters and credit costs on this page are read from the live catalog and cannot drift from what the API accepts. Field notes were captured on 2026-08-30.