Megalania
Web scraping, crawling and page monitoring for AI clients via one MCP endpoint
We have not yet received an answer to our tool-list check (latest attempt 6 hours ago).
Megalania is an MCP server for web search and scraping. It runs as a hosted remote server, so there is nothing to install locally.
- Platforms
- API, CLI, Web
- Runs
- Remotely
- Version
- 1.8.0
- Listing updated
Similar MCP servers
Install
Run in a terminal
claude mcp add --transport http megalania https://mcp.megalania.io/Add to Settings → Connectors → Add custom connector
https://mcp.megalania.io/Add to ~/.cursor/mcp.json
{
"mcpServers": {
"megalania": {
"url": "https://mcp.megalania.io/"
}
}
}Add to .vscode/mcp.json
{
"servers": {
"megalania": {
"type": "http",
"url": "https://mcp.megalania.io/"
}
}
}Add to ~/.codeium/windsurf/mcp_config.json
{
"mcpServers": {
"megalania": {
"serverUrl": "https://mcp.megalania.io/"
}
}
}Add to Grok Bot desktop app → Connect apps → Add a custom connector
https://mcp.megalania.io/Enter a Name (for example "Megalania"), choose Type: Remote HTTPS, paste this as the Server URL and choose Add plugin.
Add to your client's MCP settings (most use this shape)
{
"mcpServers": {
"megalania": {
"type": "http",
"url": "https://mcp.megalania.io/"
}
}
}Using several servers? Build one config file with the MCP config generator.
About Megalania
The Megalania MCP server gives an AI client a set of fifteen tools for reading public web pages. Pages can come back as Markdown, text, HTML, links or metadata, and a separate extract tool pulls named fields out of a page using CSS rules. The README-style docs describe a free check_site call that reports expected cost and success rates for a host before any credits are spent.
Beyond single fetches, the server runs batches of up to 10,000 pages or a crawl of one site in the background, and it can map a site's URLs without fetching their content. Monitor tools schedule checks on a set of pages and report what changed between runs. Blocked pages and errors are described as free, and the docs say the answer includes next_step guidance for the client to follow.
The server is aimed at developers and agent builders who connect Claude, ChatGPT, Cursor or VS Code to live web data. Sign-in is through a browser for Claude and ChatGPT, or a bearer key for other clients. The docs note that access requires a beta invite.
Tools this server provides (15)
Taken from the project documentation; the running server may differ.
- scrape
- Fetch one page and return it in the requested formats.
- inspect
- Return a page's structure so extraction rules can be chosen.
- extract
- Read fields from a page with CSS rules or a JSON schema.
- check_site
- Show expected cost and success rates for a host without charging.
- map_site
- List a site's URLs without fetching their content.
- create_batch
- Start a background batch of pages or a site crawl.
- get_batch
- Read batch status, items and stored page content.
- cancel_batch
- Stop a running batch.
- list_batches
- List batches, optionally filtered by status.
- get_usage
- Show credit balance and per-site usage.
- create_monitor
- Create a scheduled check on one to 100 pages.
- list_monitors
- List monitors with status and monthly credit use.
- get_monitor_changes
- Return recent checks and the newest change diffs.
- update_monitor
- Change a monitor's settings, pause it, or run it now.
- delete_monitor
- Delete a monitor and its change history.
What it does
- Multi-format page reading
- Fetches a page and returns it as Markdown by default, with text, HTML, links, metadata or structured page data available as alternatives.
- Field extraction with CSS rules
- The extract tool reads named fields and lists from a page, or fills a JSON schema using a language model.
- Structure inspection
- The inspect tool returns repeating blocks and suggested extraction rules so an agent can pick fields without reading raw HTML.
- Batches and site crawls
- create_batch runs up to 10,000 pages or a depth-limited crawl in the background, with status polling and cancellation.
- Scheduled page monitoring
- Monitors check pages on an interval or cron schedule and return diffs of content or extracted fields.
- Browser rendering for JavaScript pages
- An optional browser engine renders pages that need JavaScript, at a higher price than plain HTTP fetches.
- Cost preview before fetching
- check_site reports expected per-route prices and recent success rates for a host without charging credits.
Good for
- Letting an AI assistant read a product or article page during a research session
- Pulling price and title fields from a list of catalogue pages into a spreadsheet-ready format
- Crawling a documentation site to build a searchable corpus for an agent
- Watching a competitor's pricing page and receiving a diff when it changes
- Mapping a website's URLs before choosing which pages to scrape
Megalania FAQ
How do I connect Megalania to Claude Code?
The docs say to run claude mcp add with the http transport pointed at https://mcp.megalania.io, then sign in through the browser opened by /mcp. A headless setup uses a bearer key in the Authorization header.
Are blocked pages charged?
The docs state that blocked pages and errors are free, and that a needs_js answer is free unless HTTP was forced.
How many tools does the server expose?
The docs list fifteen tools: scrape, inspect, extract, check_site, map_site, create_batch, get_batch, cancel_batch, list_batches, get_usage and five monitor tools.
Is access open to everyone?
The docs say users must apply for the beta first and sign-in works once an invite arrives.
How do I add Megalania to Claude Code?
Run this command in your terminal: claude mcp add --transport http megalania https://mcp.megalania.io/
What are the best alternatives to Megalania?
The closest alternatives to Megalania are Scrapegraph MCP, Scrapling MCP Server, Parallel Search MCP, PyreCrawl and Firecrawl MCP Server.
Where this comes from
- Description: written from the project's website and its MCP Registry entry.
- Status: checked automatically by AIPoweredBot on 21 November 2026.
- Install steps: from the official MCP Registry entry.

