Social & content scraper APIs
Get structured social & content data through REST or MCP while Upscrape maintains the scrapers. Every endpoint with its price.
14 platforms · 160 endpoints · prices in credits, charged only when a job completes
-
POST/execute linkedin.jobs.searchweb.extractOne request shape, every endpoint
The endpoint ID and an input that follows its published schema, over REST or MCP.
Quickstart -
1 credit shown before every runThe price before you run
Most endpoints cost 1 credit. A few are priced by what they return, and the range is published.
How credits work -
failed 0 credits a job that does not completeFailed calls are free
Nothing is charged for a job that fails, and the same request again soon after is served from cache at no cost.
Credits and limits -
running completed your request never changesKept working by us
When a platform changes its pages, we fix the endpoint. Your code stays the same.
Connect an agent over MCP
Nothing matches. Try a platform name, an endpoint like reviews or jobs, or ask us to add it.
Not in the catalog? Scrape any website
Send any public page: name the fields you want, get the page as Markdown, or crawl a whole site. 1 credit a call, the same API.
Before you pick an endpoint
The short answers. The docs have the long ones.
Which platforms can I scrape with Upscrape?
Every platform on this page, each with its own endpoints, plus any public website through our website tools (extract, Markdown, crawl, sitemap, screenshot). New platforms are added to the catalog the day they ship; ask us if yours is missing.
How much does a call cost?
Each endpoint shows its price in credits before it runs; most cost 1 credit. You buy credits once, with no subscription. A failed call costs nothing, and paid credits never expire.
How do I call an endpoint?
One request shape for every endpoint: POST /execute with the endpoint ID and an input that follows its published schema, over REST or MCP. The response is JSON with the data and the credits charged.
Can an AI agent use these endpoints?
Yes. Connect any MCP client to the Upscrape MCP server with OAuth or an API key. The agent finds the endpoint, checks its price and runs it with the same credits.
Is scraping public data legal?
Upscrape reads only what a platform shows to any visitor; no platform needs your login. How you use the data is yours to decide under the laws that apply to you. We sign a data processing agreement for your legal review.
What if a platform I need is not here?
Scrape it as any website, or ask us to add the platform. We scope it as a custom project, and on Enterprise it can be a private scraper only you can use.
Social & content scraper APIs
14 platformsAre.na
9 endpointsSearch and inspect public Are.na profiles, channels, blocks, and their connections.
Bluesky
12 endpointsRead Bluesky actor and post data from public AT Protocol endpoints.
Facebook
17 endpointsExtract public Facebook pages, content, engagement metrics, and Ads Library results.
Instagram
19 endpointsPublic Instagram profiles, media, comments, reels, audio, embeds, and discovery.
LinkedIn
14 endpointsPublic LinkedIn profiles, organizations, content, jobs, newsletters, and Learning data.
Linktree
5 endpointsPublic Linktree profiles, links, creator posts, Shop products, and directory discovery.
Pinterest
9 endpointsExtract public Pinterest profiles, boards, sections, and pin metadata.
Product Hunt
9 endpointsRead Product Hunt discovery feeds, topics, categories and reviews.
Reddit
9 endpointsMonitor Reddit communities, posts, threads, and public user activity without managing browser sessions.
Threads
9 endpointsExtract public Threads profiles, tag metrics, search results, timelines, posts, and replies.
TikTok
22 endpointsPublic TikTok profiles, videos, comments, discovery, native captions, LIVE metadata and US Shop products over HTTP.
Tumblr
11 endpointsPublic Tumblr profiles, posts, reblog trails, polls, tag discovery, and original media.
X (Twitter)
1 endpointYouTube
14 endpointsExtract public YouTube videos, channels, Shorts, live broadcasts, comments, playlists, and search results.