SAVE 20%on any yearly plan

The web scraper API yourAI agent deserves.

Clean markdown or JSON from any URL, with 90% fewer tokens than raw HTML. A drop-in Firecrawl replacement, no headless browser.

3/3

Live, no signup — 3 free runs a day. The hosted API is a paid plan from $19/mo; the open-source core is free to self-host.

Want every option and no cap? Open the full playground

Cancel anytimeSelf-host foreverOpen source core

Production-ready on real-world sites like these

NikeAirbnbEtsyShopifyIKEATargetTripAdvisorZillowIMDbRotten TomatoesYelpGoodreadsWikipediaGitHubProduct HuntY CombinatorStripeSpotifyNikeAirbnbEtsyShopifyIKEATargetTripAdvisorZillowIMDbRotten TomatoesYelpGoodreadsWikipediaGitHubProduct HuntY CombinatorStripeSpotify
Before vs After Webclaw

Stop building your own scraper. Call one web scraper API instead.

Rolling your own scraper

You wire up parsers, headless browsers, and proxy pools, then babysit the stack every time a site ships a redesign and your selectors snap.
Your requests hit a bot wall, come back as an empty shell of HTML, and your agent quietly reasons over nothing.
You dump raw page markup into the model and watch nav bars, cookie banners, and script tags eat the context window.
The model answers from frozen training data, so it invents prices, docs, and news it has no way to actually read.
One page is JavaScript, the next is a PDF, a third is a DOCX, so you bolt on a different library for every format and glue them by hand.

With Webclaw

Webclaw is a web scraper API for AI agents: one call takes a URL and returns clean markdown, structured JSON, or LLM-ready text, and site redesigns become our problem, not yours.
Webclaw gets through bot walls and renders JavaScript pages, so you scrape without getting blocked and your agent reads the real content.
Webclaw converts HTML to clean markdown and strips the page down to meaning, cutting the tokens you feed a model by around 90% and leaving room in the context window for reasoning.
Webclaw feeds agents live web data on demand, so answers cite the page as it reads right now instead of stale training data.
One API turns HTML, PDFs, and DOCX into structured data across 14 REST endpoints, and its 12-tool MCP server plugs straight into Claude, Cursor, and Codex.

Webclaw is a Firecrawl alternative with a drop-in /v2 route and an open-source Rust core: swap one base URL, keep every call you already wrote, and ship product instead of scraping infra.

Product

Everything an agent needs. Nothing it doesn’t.

Fast by default, smart when needed.

Static pages come back instantly. JS-heavy sites only render when they have to, with nothing to configure.

A drop-in Firecrawl replacement.

Swap one base URL, keep your SDK. /v2 matches Firecrawl's shape and response exactly, no rewrite.

Best-in-class bot protection.

Challenge pages, CAPTCHAs and fingerprinting handled transparently. No cookies, no per-site config.

Every format, every extraction.

Markdown, JSON, text, LLM-optimized. Schema & prompt extraction, summaries, diffing, across 14 endpoints.

Built for AI agents.

An MCP server with 12 tools for Claude, Cursor, Codex & any client. REST for search, batch and crawl.

90% fewer tokens.

A 9-step pipeline strips the nav, ads and boilerplate, keeping the real content fully intact.

Agentic scraping.

Give it a goal, get structured data back. The agent reasons over the page, clicks and navigates to it.

Deep content recovery.

Embedded JSON & server-rendered payloads recovered even from an empty DOM. Auto-detects PDF, DOCX, XLSX.

0k
pages extracted
0
bot walls bypassed
0
websites scraped
0
github stars
Pricing

Pay for pages, not seats.

One credit pool covers every endpoint. Start in minutes, or self-host the open-source stack with no limits.

Save 20%
Starter
$19/mo
  • Credits 10,000/mo
  • Research 3 runs/mo
  • Max sources 10
  • Concurrency 5
  • Support Email

Card required · cancel anytime

PopularGrowth
$49/mo
  • Credits 100,000/mo
  • Research 10 runs/mo
  • Max sources 20
  • Concurrency 20
  • Support Priority
Pro
$99/mo
  • Credits 250,000/mo
  • Research 20 runs/mo
  • Max sources 30
  • Concurrency 50
  • Support Priority
Scale
$399/mo
  • Credits 1,000,000/mo
  • Research 60 runs/mo
  • Max sources 100
  • Concurrency 100
  • Support Priority + Slack

Cancel anytime·no lock-in·self-host the open-source core for free

See how credits work
Common questions

Questions, answered.

Webclaw is a web extraction toolkit that turns any website into clean, structured data. Output formats include Markdown, JSON, HTML, plain text, and an LLM-optimized mode that strips noise and cuts token count by around 90% vs raw HTML.

Webclaw uses HTTP with TLS fingerprint impersonation instead of spinning up a headless browser. Sub-200ms response times, zero browser overhead, no Selenium or Playwright dependency. Content extraction runs via readability scoring plus a 9-step pipeline, no browser needed for most pages.

The open-source version (AGPL-3.0) runs locally on your own hardware with no limits, so you can try the full engine for free without a card. The managed API is paid, starting at $19/mo for Starter, and you can cancel any time from the billing portal.

Yes. Webclaw is open source under AGPL-3.0. You can run the CLI, REST API server, or MCP server on your own infrastructure. Docker images and one-line deploy scripts are available.

Six formats: Markdown (clean readable text), JSON (structured with metadata), HTML (sanitized), plain text, LLM-optimized (stripped of noise for AI consumption), and raw HTML. The LLM format runs a 9-step optimization pipeline to minimize token usage.

Webclaw ships a Model Context Protocol server binary that exposes 12 tools: scrape, crawl, map, batch, extract, summarize, diff, brand, search, research, vertical_scrape, and list_extractors. Works with any MCP client (Claude Desktop, Claude Code, Cursor, Windsurf, Codex, Antigravity) over stdio.

Your extracted content is never stored or logged on our servers. Requests are processed in real-time and the response is returned directly to you. If you use LLM features, content is sent to the AI provider for processing but is not retained. For full control, self-host the entire stack.

Webclaw can use language models to extract structured JSON from pages using a schema you define, answer questions about page content with prompt-based extraction, or generate summaries. It chains through local Ollama first, then falls back to cloud providers.

Partners

Backing open web extraction

ColdProxyNodeMavenRapidProxyMangoProxyColdProxyNodeMavenRapidProxyMangoProxyColdProxyNodeMavenRapidProxyMangoProxyColdProxyNodeMavenRapidProxyMangoProxy
Become a partner

Ship your agent today. Scrape forever.

Cancel anytime. Migrate from Firecrawl in 60 seconds with the compatibility layer.

Star on GitHub