Open API

Open API.

The same data that feeds Open Current, free for anyone to build on. Read-only JSON, open CORS, no key, no sign-up. Every response is served from Netlify's cache, so it's fast for you and costs the project almost nothing.

Fair use & attribution

Link back to the original source. Every item has a u (URL) and s (source name). Show them, and send readers to the original. Please also mention “via Open Current” with a link to https://open-current-feed.netlify.app.
Be cache-friendly. Feeds refresh every 10 to 60 minutes (see each endpoint). Polling more often than every 5 minutes gets you the same cached answer. Extra query parameters are ignored by the cache, so don't try to bust it.
Headlines and snippets only. We pass on titles, short descriptions and links from public feeds. The articles belong to their publishers.
No personal data. We don't log requests, IPs or queries, and there are no keys to track you with.
Best effort, no guarantees: a source can go quiet or change its feed at any time. Please don't use this for anything safety-critical or financial decisions.

Endpoints

GET/api/feed/{pack}

Newest headlines from an allowlisted, https-only set of feeds, deduplicated and sorted newest first. Each sources entry is the number of items that feed returned this refresh.

pack	what	refresh
`global`	World desks: 100+ newsrooms, 30+ languages, every region (items carry `l` language and `r` region)	30 min
`world`	Major English-language world news	30 min
`techpress`	Tech press	30 min
`war`	Conflict and humanitarian reporting	20 min
`memes`	Meme communities (image posts)	20 min
`arxiv`	New arXiv papers	2 h
`ytchannels`	New videos from public YouTube channels	60 min
`substack`	Independent newsletters	2 h
`art`	Art and design publications, NFT/crypto-art news and Lemmy art communities (NSFW and nudity-worded items removed)	30 min

/api/feed/global /api/pulse /api/feed/globalplus /api/feed/world /api/feed/arxiv

GET/api/crypto-news

Headlines from 13 crypto newsrooms, newest first. Refreshes every 20 minutes.

Try it

GET/api/discover?q={query}

Independent-web search across Wiby, arXiv and Kagi Small Web, with big platforms, insecure and risky links removed. Cached for an hour per query. ?surprise=1 returns a list of random small-web pages. (On the Discover page, Marginalia, Hacker News, Lemmy, Mastodon, OpenAlex and Wikipedia are asked directly from the browser.)

?q=sourdough

Item format

{
  "updated": 1794579407000,          // when this response was built (ms since 1970, UTC)
  "sources": { "DW English": 3, … },  // items per feed in this refresh
  "items": [{
    "s":  "RTP Notícias",             // source / outlet name
    "t":  "Headline",                 // title
    "u":  "https://…",                // original link (https only, checked)
    "d":  "Short description",        // snippet, may be empty
    "i":  "https://…jpg",             // image, optional
    "ts": 1794579407000,              // published time (ms, UTC; never in the future)
    "l":  "pt",                       // language, when known
    "r":  "europe",                   // region (global pack)
    "k":  ["news"]                    // topics
  }]
}

Discover returns { q, results:[{ u, t, d, e (engine), c (category) }], counts, dropped, errors }.

Live streams: connect directly

Open Current doesn't proxy the real-time firehoses: your app can connect to them straight away, exactly like our pages do in your browser.

WSSBluesky Jetstream

Every public Bluesky post as JSON. Add &cursor= (microseconds) to replay up to about a day.

wss://jetstream2.us-east.bsky.network/subscribe?wantedCollections=app.bsky.feed.post

WSSNostr relays

Send ["REQ","sub",{"kinds":[1],"since":<unix seconds>}] to any public relay.

wss://relay.primal.net  ·  wss://nos.lol  ·  wss://offchain.pub  ·  wss://relay.nostr.net

SSEWikimedia EventStreams

Server-sent events for every new page, edit and upload across Wikimedia projects.

https://stream.wikimedia.org/v2/stream/page-create
https://stream.wikimedia.org/v2/stream/recentchange

GETAlso public

Hacker News: https://hacker-news.firebaseio.com/v0/newstories.json · Mastodon: https://{server}/api/v1/timelines/public?local=true · Lemmy: https://{server}/api/v3/post/list?sort=New · Misskey: POST https://{server}/api/notes/local-timeline. Please respect each network's own rate limits and terms.

Example

const r = await fetch("https://open-current-feed.netlify.app/api/feed/global");
const { items } = await r.json();
for (const it of items.slice(0, 10))
  console.log(`${it.t} (${it.s}) ${it.u}`);   // always keep the source link

For AI tools and crawlers there's a plain-text guide at /llms.txt. Questions or a feed to suggest? See Sources.