# scrapeless-ai/scrapeless-mcp-server [Health: Active]

**Category:** 🔎 Search & Data Extraction  
**Repository:** https://github.com/scrapeless-ai/scrapeless-mcp-server  
**GitHub Stars:** 168  
**npm Downloads (last month):** 1176  
**Views:** 3  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/scrapeless-ai-scrapeless-mcp-server

## Description
The Scrapeless Model Context Protocol service acts as an MCP server connector to the Google SERP API, enabling web search within the MCP ecosystem without leaving it.

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `npx` (confidence: high):

```json
"mcpServers": {
  "scrapeless-mcp-server": {
    "command": "npx",
    "args": ["-y","scrapeless-mcp-server"],
    "env": {
      "SCRAPELESS_API_KEY": "",
      "BROWSER_PROFILE_ID": "",
      "BROWSER_PROFILE_PERSIST": "",
      "BROWSER_SESSION_TTL": ""
    }
  }
}
```

**Requires environment variables:** `SCRAPELESS_API_KEY`, `BROWSER_PROFILE_ID`, `BROWSER_PROFILE_PERSIST`, `BROWSER_SESSION_TTL` — the values above are empty placeholders; fill in real credentials before running (see the repository for what each one is for).

## Documentation

## What scrapeless-ai/scrapeless-mcp-server MCP server does

The scrapeless-ai/scrapeless-mcp-server MCP server gives compatible AI clients access to Scrapeless web capabilities through the Model Context Protocol. Its documented scope covers Google Search, Google Trends, cloud browser sessions, page navigation, interaction, and content extraction. Agents can request visible page text, full HTML, screenshots, and browser-based actions such as clicking, typing, scrolling, waiting, and pressing keys.

The README also describes scraping JavaScript-heavy pages into HTML, Markdown, or screenshots, crawling websites by following links, and creating AI Scraper tasks for several named AI search and assistant services. These broader functions are part of the Scrapeless service exposed through the MCP integration; the excerpted tool table does not list every related tool.

## How it works

The server supports two transport options. In stdio mode, an MCP client launches the npm package with `npx` and passes the Scrapeless credential through the `SCRAPELESS_API_KEY` environment variable. In Streamable HTTP mode, the client connects to `https://api.scrapeless.com/mcp` and sends the key in the `x-api-token` header.

A typical browser workflow starts by creating or reusing a cloud browser session, then navigating to a URL and performing actions against the page. The agent can wait for elements or fixed periods, retrieve rendered content, capture a screenshot, and close the session when finished. Browser sessions can use a reusable profile and optional persistence for cookies and local storage.

## Setup and configuration

Create a Scrapeless API key from the Scrapeless Dashboard under Setting, API Key Management, and Create API Key. The project states that a free trial is available. Add the key to the client configuration before enabling the server.

For local execution, the documented configuration uses:

```json
{
  "command": "npx",
  "args": ["-y", "scrapeless-mcp-server"],
  "env": {
    "SCRAPELESS_API_KEY": "your-key"
  }
}
```

The hosted configuration uses the Streamable HTTP URL and an `x-api-token` header. Optional browser settings are available as environment variables in stdio mode or matching HTTP headers remotely:

- `BROWSER_PROFILE_ID` selects a reusable browser profile.
- `BROWSER_PROFILE_PERSIST` enables persistent browser storage.
- `BROWSER_SESSION_TTL` sets the inactivity timeout in seconds.

## Tools and capabilities

The documented tools include:

- `google_search` for general Google information searches.
- `google_trends` for Google Trends data.
- `browser_create` and `browser_close` for cloud browser session lifecycle.
- `browser_goto`, `browser_go_back`, and `browser_go_forward` for navigation.
- `browser_click`, `browser_type`, and `browser_press_key` for page interaction.
- `browser_wait_for` and `browser_wait` for synchronization.
- `browser_screenshot`, `browser_get_html`, and `browser_get_text` for page output.
- `browser_scroll` for moving through a page.

The scrapeless-ai/scrapeless-mcp-server MCP server is therefore suited to research, search-result collection, rendered-page extraction, and agent tasks that require browser interaction rather than only HTTP text retrieval.

## Limitations and notes

A Scrapeless API key is required, so usage depends on the Scrapeless service and its account terms. The excerpt does not provide a complete tool list, pricing details beyond a free-trial reference, request limits, or supported output schemas. Browser profile settings are optional, and session timeout behavior is controlled through the configured TTL. Client setup differs between local stdio and hosted Streamable HTTP, so choose the transport supported by the MCP client you are configuring.

_Full upstream README: https://allmcps.com/mcp/scrapeless-ai-scrapeless-mcp-server/readme_