Bluesky, Mastodon, Lemmy, Discourse and Tumblr

Posts from social networks and forums as typed JSON, every one in the same post shape, so feeds merge without mapping.

Embed

<a href="https://scrapelyserp.com/docs/social"><img src="https://scrapelyserp.com/docs/_api/public/orgs/affc5b0f-d3c2-423a-8143-9d10db5736ae/spaces/981e6275-da3f-446c-82de-889f59c83b6f/pages/248d6a87-72eb-45fe-af62-56091cf272de/social-card.svg?v=3e7ae33eec8b" alt="Bluesky, Mastodon, Lemmy, Discourse and Tumblr" width="600" height="315"></a>

Posts from social networks and forums as typed JSON, every one in the same post shape, so feeds merge without mapping.

At a glance

Every endpoint here answers { endpoint, url, credits, duration_ms, data } with data as a SocialFeed: posts and next_cursor. Each post has the same fields whatever the site: id, url, author, author_name, title, text, created_at, likes, replies, reposts, tags, community, link and media. A count a site does not show is null. Send next_cursor back as cursor for the next page.

Endpoint	Typical
`GET /v1/structured/bluesky/search`	about 4.8 million posts per credit, up to 100 per call
`GET /v1/structured/bluesky/posts`	about 1.4 million posts per credit, up to 100 per call
`GET /v1/structured/mastodon/hashtag`	about 2.8 million posts per credit, up to 40 per call
`GET /v1/structured/lemmy/search`	about 2.3 million posts per credit, up to 50 per call
`GET /v1/structured/lemmy/posts`	about 3.5 million posts per credit, up to 50 per call
`GET /v1/structured/discourse/search`	about 7.5 million posts per credit, 50 per call
`GET /v1/structured/discourse/topic`	about 750,000 posts per credit, 20 per call
`GET /v1/structured/tumblr/tagged`	about 450,000 posts per credit, up to 20 per call
`GET /v1/structured/tumblr/search`	about 450,000 posts per credit, up to 20 per call

cURL
Python
Node.js

curl -G "https://api.scrapelyserp.com/v1/structured/bluesky/search" \
  -H "Authorization: Bearer $SCRAPELY_API_KEY" \
  --data-urlencode "query=web scraping" \
  -d sort=latest

import os
import requests

HEADERS = {"Authorization": f"Bearer {os.environ['SCRAPELY_API_KEY']}"}

r = requests.get(
    "https://api.scrapelyserp.com/v1/structured/bluesky/search",
    headers=HEADERS,
    params={"query": "web scraping", "sort": "latest"},
    timeout=30,
)
r.raise_for_status()
data = r.json()
for p in data["data"]["posts"]:
    print(p["created_at"], p["author"], p["text"][:80])

const res = await fetch(`https://api.scrapelyserp.com/v1/structured/bluesky/search?${new URLSearchParams({ query: 'web scraping', sort: 'latest' })}`, {
  headers: { Authorization: `Bearer ${process.env.SCRAPELY_API_KEY}` },
  signal: AbortSignal.timeout(30_000),
});
if (!res.ok) throw new Error((await res.json()).error.code);
const data = await res.json();
for (const p of data.data.posts) console.log(p.created_at, p.author, p.text.slice(0, 80));

Watching for mentions

A mention list does it for you: name a brand or a phrase, and every run searches Reddit, Bluesky, Lemmy and Tumblr newest first and sends each post you have not seen as watchlist.mention.found.

A handle, community, topic or server that does not exist (or is not public) is 404 NOT_FOUND, and a page we could not read is 502 TARGET_FAILED. Neither costs anything.

Bluesky

bluesky/search searches every public post; bluesky/posts is one account's posts, newest first, without its reposts of others.

Parameter	Type	Default	Description
`query` (required)	string	—	The search terms.
`sort`	`latest`, `top`	`latest`	Result order.
`limit`	integer	`25`	Posts per page, 1–100.
`cursor`	string	—	The `next_cursor` of the previous page.

Parameter	Type	Default	Description
`handle` (required)	string	—	The account, e.g. `bsky.app` (also `@bsky.app` or its profile URL).
`limit`	integer	`25`	Posts per page, 1–100.
`cursor`	string	—	The `next_cursor` of the previous page.

Full reference → GET /v1/structured/bluesky/search, GET /v1/structured/bluesky/posts

Mastodon

Public posts with a hashtag, on mastodon.social or any server you name with instance. A server also shows posts from the servers it federates with, so author always carries its server: name@server. reposts counts boosts.

Parameter	Type	Default	Description
`tag` (required)	string	—	The hashtag, with or without `#`, e.g. `webscraping`.
`instance`	string	`mastodon.social`	The Mastodon server to read, e.g. `fosstodon.org`. It also shows posts from servers it federates with.
`limit`	integer	`20`	Posts per page, 1–40.
`cursor`	string	—	The `next_cursor` of the previous page.

Full reference → GET /v1/structured/mastodon/hashtag

Lemmy

lemmy/search searches posts; lemmy/posts is one community's. likes is the score (up votes less down votes), replies the comment count, community is name@server.

Parameter	Type	Default	Description
`query` (required)	string	—	The search terms.
`instance`	string	`lemmy.world`	The Lemmy server to read, e.g. `lemmy.ml`. It also has the communities it federates with.
`sort`	enum	`new`	Result order.
`limit`	integer	`20`	Posts per page, 1–50.
`cursor`	string	—	The `next_cursor` of the previous page (a page number).

Parameter	Type	Default	Description
`community` (required)	string	—	The community, e.g. `technology`, or one on another server: `technology@lemmy.ml`.
`instance`	string	`lemmy.world`	The Lemmy server to read, e.g. `lemmy.ml`. It also has the communities it federates with.
`sort`	enum	`new`	Result order.
`limit`	integer	`20`	Posts per page, 1–50.
`cursor`	string	—	The `next_cursor` of the previous page (a page number).

Full reference → GET /v1/structured/lemmy/search, GET /v1/structured/lemmy/posts

Discourse forums

Any public Discourse forum: name it with forum. discourse/search returns matching posts with their topic's title, tags and category (community), and an excerpt as text. discourse/topic returns a topic's posts in order, full text, the opening post first. A forum that needs a login is 404 NOT_FOUND.

Parameter	Type	Default	Description
`forum` (required)	string	—	The forum’s domain, e.g. `meta.discourse.org` (or any URL on it).
`query` (required)	string	—	The search terms.
`cursor`	string	—	The `next_cursor` of the previous page (a page number).

Parameter	Type	Default	Description
`forum` (required)	string	—	The forum’s domain, e.g. `meta.discourse.org` (or any URL on it).
`topic` (required)	string	—	The topic’s id, e.g. `49045`, or its URL.
`cursor`	string	—	The `next_cursor` of the previous page (a page number).

Full reference → GET /v1/structured/discourse/search, GET /v1/structured/discourse/topic

Tumblr

tumblr/tagged is posts with a tag; tumblr/search searches by keyword. author is the blog, likes the likes (or notes), reposts the reblogs. Promoted posts are left out.

Parameter	Type	Default	Description
`tag` (required)	string	—	The tag, e.g. `webscraping`.
`sort`	`recent`, `top`	`recent`	Result order.
`limit`	integer	`20`	Posts per page, 1–20.
`cursor`	string	—	The `next_cursor` of the previous page.

Parameter	Type	Default	Description
`query` (required)	string	—	The search terms.
`sort`	`recent`, `top`	`recent`	Result order.
`limit`	integer	`20`	Posts per page, 1–20.
`cursor`	string	—	The `next_cursor` of the previous page.

Full reference → GET /v1/structured/tumblr/tagged, GET /v1/structured/tumblr/search

Updated Nov 6, 2026

Was this page helpful?