Migrate from Firecrawl to WebPeel in 5 Minutes

WebPeel is a free, open-source alternative to Firecrawl with a compatible API. In most cases, switching is a one-line URL change.

Why switch?

Firecrawl is a well-built product and a solid choice. If you're on their free tier and it works for your needs, there's no urgent reason to switch. WebPeel is for people who want:

• A free recurring tier — 500 requests/week, no credit card, no one-time cap
• Self-hosting without restrictions — full AGPL-3.0 codebase, docker compose up and you're running
• More MCP tools — 18 vs. ~5 (quick answers, YouTube transcripts, hotel search, change tracking, and more)
• Stealth mode built-in — 28 preconfigured domains with anti-bot bypass, no cloud-only restriction
• LLM-free Q&A — answer factual questions from web pages without sending content to an LLM
• Token efficiency — budget mode reduces tokens by 83% on typical pages

Feature comparison

API compatibility

WebPeel supports Firecrawl-compatible endpoints at /v1/scrape and /v1/crawl. For most use cases, switching is a single URL change:

Supported Firecrawl-compatible fields:

• POST /v1/scrape — url, formats (markdown, html, rawHtml, screenshot, links, json), onlyMainContent, waitFor, actions
• POST /v1/crawl — url, limit, maxDepth, includePaths, excludePaths
• GET /v1/crawl/:jobId — job status polling
• POST /v2/scrape — v2 scrape endpoint

Code examples

REST API

curl -X POST https://api.firecrawl.dev/v1/scrape \
  -H "Authorization: Bearer fc-YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "url": "https://example.com",
    "formats": ["markdown"]
  }'

curl -X POST https://api.webpeel.dev/v1/scrape \
  -H "Authorization: Bearer wp_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "url": "https://example.com",
    "formats": ["markdown"]
  }'

The response shape is identical. No other changes needed for basic scraping.

MCP server config (Claude Desktop)

{
  "mcpServers": {
    "firecrawl": {
      "command": "npx",
      "args": ["-y", "firecrawl-mcp"],
      "env": {
        "FIRECRAWL_API_KEY": "fc-YOUR_KEY"
      }
    }
  }
}

{
  "mcpServers": {
    "webpeel": {
      "command": "npx",
      "args": ["-y", "webpeel", "mcp"]
    }
  }
}

No API key is required to use the WebPeel MCP server. You get 25 anonymous requests per hour, or sign up free for 500/week.

Prefer a remote URL (no install, always up-to-date)?

{
  "mcpServers": {
    "webpeel": {
      "url": "https://api.webpeel.dev/v2/mcp"
    }
  }
}

Node.js library

import FirecrawlApp from '@mendable/firecrawl-js';

const app = new FirecrawlApp({
  apiKey: 'fc-YOUR_KEY'
});

const result = await app.scrapeUrl(
  'https://example.com',
  { formats: ['markdown'] }
);

console.log(result.markdown);

import { peel } from 'webpeel';

// No API key required for basic usage
const result = await peel('https://example.com');

console.log(result.content); // markdown

Install WebPeel: npm install webpeel

For API-key authenticated requests with the full client:

import { WebPeel } from 'webpeel';

const client = new WebPeel({ apiKey: 'wp_YOUR_KEY' });

const result = await client.scrape('https://example.com', {
  formats: ['markdown'],
  budget: 4000,     // Token budget (WebPeel-specific)
  stealth: true,    // Built-in stealth mode
});

console.log(result.markdown);

Python SDK

from firecrawl import FirecrawlApp

app = FirecrawlApp(api_key="fc-YOUR_KEY")

result = app.scrape_url(
    "https://example.com",
    params={"formats": ["markdown"]}
)

print(result["markdown"])

from webpeel import WebPeel

client = WebPeel(api_key="wp_YOUR_KEY")

result = client.scrape("https://example.com")

print(result.content)  # markdown

Install: pip install webpeel

What's different (and new)

WebPeel does everything Firecrawl does, plus a few things it doesn't:

Budget mode — token-efficient fetching

Pass budget=4000 to get any page distilled to ~4K tokens. Content density pruning removes sidebars, footers, navigation, and ads automatically. On a Wikipedia page, this reduced 99,427 tokens to 3,867 — a 96% reduction.

# REST API
curl "https://api.webpeel.dev/v1/fetch?url=https://en.wikipedia.org/wiki/AI&budget=4000"

# CLI
npx webpeel https://en.wikipedia.org/wiki/AI --budget 4000

LLM-free quick answers

Ask factual questions from any web page without sending content to an LLM. Uses BM25 retrieval + heuristic answer extraction with a confidence score.

curl -X POST https://api.webpeel.dev/v1/answer \
  -H "Content-Type: application/json" \
  -d '{
    "url": "https://example.com/pricing",
    "question": "What is the monthly price?"
  }'

Domain extractors (stealth built-in)

28 domain-specific extractors handle sites that typically require anti-bot bypasses: Amazon, Reddit, LinkedIn, Twitter/X, GitHub, Hacker News, Yelp, Booking.com, and more. These work on self-hosted instances too — no cloud requirement.

npx webpeel https://www.amazon.com/dp/B08N5WRWNW --schema amazon --json

YouTube transcripts

Extract full transcripts from YouTube videos — no YouTube API key required.

npx webpeel https://www.youtube.com/watch?v=dQw4w9WgXcQ

Hotel search

Multi-source hotel search with merged and sorted results — a unique capability not available in other web fetching tools.

npx webpeel hotels "Paris" --checkin 2026-04-01 --checkout 2026-04-05 --sort price

Step-by-step migration

npm install webpeel       # Node.js
pip install webpeel       # Python

Self-hosting

Unlike Firecrawl, all WebPeel features (including stealth mode and domain extractors) work in self-hosted deployments.

# Clone and start
git clone https://github.com/webpeel/webpeel
cd webpeel
docker compose up

See the Self-Hosting guide for environment variables, database setup, and production configuration.

Next steps

• API Reference — full parameter docs for WebPeel's native API
• MCP Server — all 7 MCP tools documented
• LLM-Free Q&A — how the answer endpoint works
• Node.js & Python SDKs — full SDK reference
• Self-Hosting — run WebPeel on your own infrastructure
• GitHub — source code, issues, contributions