> Documentation index: https://docs.scrapier.io/llms.txt
> Auth: header `X-API-Key: <key>` · Base: https://api.scrapier.io/api/v2 · Synchronous; paginate with `cursor` ← `data.next_cursor`.

# Claude connector

Add Scrapier to Claude on claude.ai and Claude Desktop, sign in with your Scrapier account, set a daily credit limit and revoke access at any time.

The Scrapier connector lets Claude look up public web data (products, videos, listings, places, profiles) from the platforms in the [API reference](https://docs.scrapier.io/api-reference.md). You sign in with your Scrapier account, so there is no API key to copy or paste.

It is the same [MCP server](https://docs.scrapier.io/ai/mcp.md) that Claude Code, Cursor and Codex use, so it reads the same live catalog and bills the same way.

|                |                                                                                                                                                                                                           |
| -------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| **Server URL** | `https://mcp.scrapier.io/mcp`                                                                                                                                                                             |
| **Sign-in**    | Your Scrapier account (OAuth 2.1)                                                                                                                                                                         |
| **Needs**      | A Scrapier account with credits, and a Claude plan that allows custom connectors. On Team and Enterprise plans an organization Owner adds it first.                                                       |
| **Cost**       | Free to discover endpoints and check your balance. Each successful scrape costs the endpoint's `cost_credit` (1 to 4 credits, depending on the endpoint; see [Credits & pricing](https://docs.scrapier.io/credits-and-pricing.md)). |

## Add the connector

1. Open the install link: [Add Scrapier to Claude](https://claude.ai/customize/connectors?modal=add-custom-connector\&connectorName=Scrapier\&connectorUrl=https%3A%2F%2Fmcp.scrapier.io%2Fmcp). Or in Claude go to **Customize → Connectors → Add custom connector** and enter the server URL above.
2. Click **Connect** and sign in with your Scrapier account. Create one at [app.scrapier.io](https://app.scrapier.io) first if you don't have one.
3. On the Scrapier page, check the app name, the permissions and the **daily credit limit**, then click **Approve**.
4. Back in Claude, turn the connector on in the **+** menu of a chat and ask for something.

> **Team and Enterprise:** An organization Owner adds the connector once from the organization's connector settings. Each member then connects with their own Scrapier account, so each person's scrapes use their own credits.

## Try it

- "Using Scrapier, find the scraper for Amazon product pages and show me its parameters."
- "Use Scrapier to get the title, price and rating for this Amazon product: _paste a product URL_."
- "How many Scrapier credits do I have left?"

Claude searches the catalog, reads the endpoint's parameters, then runs it. Claude asks before every scrape because a scrape spends credits.

## What Claude can do

The connector has six tools. See [MCP server](https://docs.scrapier.io/ai/mcp.md#tools) for the full table.

| Tool                  | Cost                         | What it does                                       |
| --------------------- | ---------------------------- | -------------------------------------------------- |
| `list_platforms`      | free                         | Every platform and its endpoint count.             |
| `search_endpoints`    | free                         | Find endpoints by keyword.                         |
| `get_endpoint_schema` | free                         | The parameters for one endpoint.                   |
| `run_scrape`          | The endpoint's `cost_credit` | Runs one endpoint and returns one page of results. |
| `get_scrape_result`   | free                         | Kept for older clients.                            |
| `get_account`         | free                         | Your email, plan and credit balance.               |

Some slower endpoints can't run from a connector because they outlast a chat tool call. Claude is told which ones, nothing is billed for the refused call, and you can run them through the [REST API](https://docs.scrapier.io/quickstart.md) instead.

## Your control

- **Daily credit limit.** Each connection gets its own limit, 1,000 credits per UTC day by default. When it is reached, Claude's scrapes stop and nothing more is spent until 00:00 UTC. You can lower it, or raise it up to 10,000, in the dashboard. It can't be raised from inside Claude.
- **Rate limits.** 60 tool calls per minute and 600 per hour per connection, and 3 scrapes running at once across all your connectors. Claude is told when to retry.
- **Its own key.** Each connection uses a dedicated API key that only the connector holds. Revoking the connection disables that key straight away. Your other keys are untouched.
- **Revoke any time.** Open [Connected apps](https://app.scrapier.io/dashboard/connected-apps) to see each connection, its credits used today and over 30 days, and to **Revoke** it. A connection that goes unused for 30 days is revoked automatically.

## What we collect

Signing in shares your Scrapier account email with the connector so it can show your balance. A scrape sends the endpoint and inputs Claude chose to Scrapier, and the results go back to Claude in your conversation. Scrapier keeps the usual request history shown in your [History](https://app.scrapier.io/dashboard/history), plus the connection's usage totals. See the [Privacy Policy](https://scrapier.io/privacy) for retention and your rights.

## Use scraped content safely

Pages are written by whoever controls the site, not by you. Scrapier marks every result as untrusted, and Claude should read it as data, not follow instructions inside it. Check the [responsible use](https://docs.scrapier.io/responsible-use.md) page before you scrape personal data.

## Troubleshooting

| Problem                                             | Fix                                                                                                                                          |
| --------------------------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------- |
| "Couldn't connect" or the sign-in page doesn't open | Allow pop-ups for claude.ai, then retry **Connect**. If your organization restricts connectors, ask an Owner to add Scrapier.                |
| Claude says the daily limit was reached             | Wait for 00:00 UTC, or raise the limit in [Connected apps](https://app.scrapier.io/dashboard/connected-apps).                                |
| Claude says you're out of credits                   | Add credits or change plan in the [dashboard](https://app.scrapier.io/dashboard). Nothing is billed for a refused call.                      |
| The connector shows as disconnected                 | Open the connector in Claude and click **Connect** again. Sign-in tokens refresh on their own, but a revoked connection must be re-approved. |
| An endpoint is refused as too slow for a connector  | Use the [REST API](https://docs.scrapier.io/quickstart.md), or ask Claude for another endpoint.                                                                        |

Still stuck? See [Support](https://docs.scrapier.io/support.md).

- [MCP server](https://docs.scrapier.io/ai/mcp.md): Claude Code, Cursor, Codex and other clients, with an API key.
- [ChatGPT & Codex](https://docs.scrapier.io/ai/chatgpt.md): Use Scrapier from ChatGPT workspaces and Codex.
