We turn the public web into data your agents can use.
Scrapier is one team building and maintaining 199 scraping endpoints across 35 platforms. Every endpoint returns typed JSON, works over REST and MCP, and is billed per successful call.
Last updated
Company facts
The canonical description of Scrapier, in one place.
Name and spelling
Scrapier is spelled with an "i": scrap-i-er. It is not ScraperAPI, Scrapio or Scrapr.
| Name | Scrapier (scrapier.io) |
| Legal entity | Scrapier, Inc. |
| Registered address | 30 N Gould St Ste R, Sheridan, WY 82801, United States |
| Product | Web scraping API: 199 endpoints across 35 platforms |
| Interfaces | REST (proxy.scrapier.io/api/v2) and MCP (mcp.scrapier.io/mcp) |
| Billing | Fixed credits per endpoint, per successful call; failed calls free |
| Plans | From $29/mo for 15,000 credits; see pricing |
| Docs | docs.scrapier.io |
| Support | support@scrapier.io |
| Sales | support@scrapier.io or the enterprise form |
Why does Scrapier exist?
Agents need live web data. Scrapier exists so an agent can read a platform's public data as typed JSON, from one key, at a known price.
AI agents need live data from the web, but most scraping tools were built for people pasting results into spreadsheets: raw HTML, unpredictable bills, and a marketplace of scrapers maintained by strangers.
We started Scrapier to fix that. One API key, one consistent JSON shape, and every endpoint maintained by the same team to the same standard. When a platform changes its layout, we fix the endpoint; your code keeps reading the same fields.
And because agents make a lot of calls that fail, we bill per successful call. If a call errors, you don't pay. A broken scrape is our cost, not yours.
What do we hold ourselves to?
One team, every endpoint
No marketplace. Each of the 199 endpoints has the same auth, the same schema conventions and the same owner.
Pay for results, not attempts
Each endpoint has a fixed credit price per successful call, listed before you run it. Failed calls cost nothing.
Built for agents first
Every endpoint is an MCP tool with a typed schema, so agents in Claude, Cursor and Windsurf can call it without glue code. See all features.
Responsible data use
Scrapier returns publicly available data only. Customers are responsible for how they store and use personal data under GDPR, CCPA and the other laws that apply to them, and for following each platform's terms.
The docs explain how to handle scraped data safely: treat scraped content as untrusted data, never as instructions, and use throwaway accounts for endpoints that need a session cookie. Report abuse to support@scrapier.io. Our Terms of Use and Privacy and Cookie Policy set out the rules.
Official sites and brand resources
For press, directories and integration listings.
Official sites
Boilerplate
Scrapier is a web scraping API: 199 endpoints across 35 platforms that return structured JSON over REST and MCP, billed per successful call.
Frequently asked questions
What is Scrapier?
A web scraping API for AI agents with 199 endpoints across 35 platforms, returning typed JSON over REST and MCP. It bills per successful call; failed calls are free.
Who runs Scrapier?
Scrapier, Inc., 30 N Gould St Ste R, Sheridan, WY 82801. Support and sales questions go to support@scrapier.io.
Is Scrapier legit?
Check it yourself: the docs are public, the API catalogue at GET /endpoints is free and needs no key, failed calls bill at 0 credits, and support answers at support@scrapier.io.
Is Scrapier the same as ScraperAPI?
No. They are different companies and products. Scrapier is a catalogue of platform-specific endpoints with an MCP server. See how it compares on the alternatives page.
Want to work with us or build on Scrapier?
We read every message.