Get Started in 60 Seconds
Fetch your first URL, unlock your free weekly quota, connect to Claude or Cursor via MCP, and call the REST API — all in a few minutes.
Install WebPeel
curl -fsSL https://webpeel.dev/install.sh | bash
The install script adds webpeel to your PATH automatically. Restart your terminal or run source ~/.zshrc after installing.
irm https://webpeel.dev/install.ps1 | iex
Run in PowerShell (Administrator). The installer adds webpeel to your PATH automatically.
npm install -g webpeel
Works on any platform with Node.js 18+. The webpeel command is available globally after install.
# Or with pnpm / yarn
pnpm add -g webpeel
yarn global add webpeel
Verify the install:
webpeel --version
Fetch Your First URL
webpeel "https://example.com"
That's it. WebPeel outputs clean markdown — ready to paste into your AI agent, save to a file, or pipe to another command.
# Output as JSON (includes metadata)
webpeel "https://example.com" --json
# Force browser rendering for JS-heavy sites
webpeel "https://x.com/OpenAI" --render
# Use stealth mode for Cloudflare-protected sites
webpeel "https://protected-site.com" --stealth
# Fetch and save to file
webpeel "https://example.com" > output.md
--render or --stealth to force a specific mode.
Unlock 500 Free Fetches / Week
webpeel login
Follow the prompts to create a free account. You'll get:
- 500 fetches/week — resets every Monday
- 50/hour burst limit — for rapid workflows
- All features — stealth, crawl, search, MCP, research, extraction
- API key — for REST API and SDK access
After logging in, your API key is stored at ~/.webpeel/config.json and used automatically by the CLI and SDK.
Use with AI (MCP)
WebPeel includes an MCP server with 18 tools for Claude, Cursor, VS Code, and any other MCP-compatible AI client.
Add to ~/Library/Application Support/Claude/claude_desktop_config.json:
{
"mcpServers": {
"webpeel": {
"command": "webpeel",
"args": ["mcp"]
}
}
}
Restart Claude Desktop. You'll see a 🔌 icon indicating WebPeel is connected.
Open Cursor Settings → MCP → Add Server:
{
"mcpServers": {
"webpeel": {
"command": "webpeel",
"args": ["mcp"]
}
}
}
Or run from the terminal and Cursor will auto-detect the MCP server.
Add to your VS Code settings.json (with GitHub Copilot or Continue):
{
"mcp.servers": {
"webpeel": {
"command": "webpeel",
"args": ["mcp"]
}
}
}
Once connected, your AI can fetch URLs, search the web, extract structured data, get YouTube transcripts, run deep research, and more — all through natural language.
Use the REST API
Get your API key from app.webpeel.dev → Settings → API Keys.
curl -H "Authorization: Bearer YOUR_API_KEY" \
"https://api.webpeel.dev/v1/fetch?url=https://example.com"
# Search the web
curl -H "Authorization: Bearer YOUR_API_KEY" \
"https://api.webpeel.dev/v1/search?q=latest+AI+news"
# Get a YouTube transcript
curl -H "Authorization: Bearer YOUR_API_KEY" \
"https://api.webpeel.dev/v1/fetch?url=https://youtube.com/watch?v=VIDEO_ID"
import { WebPeel } from 'webpeel';
const client = new WebPeel({ apiKey: process.env.WEBPEEL_API_KEY });
// Fetch a URL
const result = await client.fetch('https://example.com');
console.log(result.content); // Clean markdown
// Search the web
const search = await client.search('latest AI news');
console.log(search.results);
// Crawl a site
const pages = await client.crawl('https://docs.example.com', {
maxPages: 50
});
from webpeel import WebPeel
client = WebPeel(api_key="YOUR_API_KEY")
# Fetch a URL
result = client.fetch("https://example.com")
print(result.content) # Clean markdown
# Search the web
search = client.search("latest AI news")
for r in search.results:
print(r.title, r.url)
# Get YouTube transcript
transcript = client.fetch("https://youtube.com/watch?v=VIDEO_ID")
print(transcript.content)
The base URL is https://api.webpeel.dev/v1. Full reference: API Reference →