---
source_url: "https://zlurp.ai/?utm_source=openai"
title: zlurp — Web Scraping API for AI Agents
mirrored_at: 2026-08-07T01:38:09.111Z
host: zlurp.ai
cited_in_42a: true
mirror_canonical: "https://index.42a.ai/zlurp.ai/index__q__utm_source_openai"
---

> **Original source:** https://zlurp.ai/?utm_source=openai

[zlurp](#)

-   [How it works](#how)
-   [API](#api)
-   [Pricing](#pricing)
-   [Compare](https://zlurp.ai/compare)
-   [Docs](https://zlurp.ai/openapi.json)
-   [Read docs →](https://zlurp.ai/openapi.json)

x402 native · no API keys required · agent-ready · USDC on Base · JS rendering available · robots.txt compliant · OpenAPI spec · llms.txt included · x402 native · no API keys required · agent-ready · USDC on Base · JS rendering available · robots.txt compliant · OpenAPI spec · llms.txt included ·

● Live — x402 v2

## Any URL.  
_Clean markdown._  
Instantly.

A web scraping API built natively for AI agents. Send a URL, receive structured markdown. Pay $0.005 per scrape via USDC — no accounts, no subscriptions, no friction.

$0.005per URL scraped

~1.2savg response time

0accounts needed

zlurp API · agent client

\# 1. Probe cost before paying

curl "https://zlurp.ai/probe?url=https://news.ycombinator.com"

→ { "costUSDC": "0.005000" }

\# 2. Scrape with x402 payment

curl \-X POST https://zlurp.ai/scrape

\-d '{"url": "https://news.ycombinator.com"}'

→ { "success": true,

"markdown": "# Hacker News ...",

"wordCount": 842 }

How it works

## Three steps. _Zero setup._

No accounts to create. No API keys to manage. Any agent with a Base wallet can start scraping in minutes.

01

🔍

### Probe the URL

Call `/probe` with any URL. Get back the exact cost before committing payment. Always free.

02

⚡

### Pay via x402

POST to `/scrape`. The x402 protocol settles $0.005 USDC on Base automatically — no human required.

03

📄

### Receive markdown

Get clean, structured markdown back. Article mode strips nav and ads. Full mode returns the entire page.

04

🤖

### Agent-native

No sessions, no credentials, no subscriptions. Any agent with a funded Base wallet can call zlurp autonomously at 3am.

API reference

## Simple endpoints. _Predictable responses._

Base URL: `https://zlurp.ai`

GET/healthno auth

Service status check. Required for AgentReady compliance.

GET/probeno auth

Cost estimate before payment. Pass `?url=` to get exact USDC cost.

POST/scrapex402 payment

Scrape any URL to markdown. Body: `{"url":"...","mode":"article"}`

GET/openapi.jsonno auth

Full OpenAPI 3.1 spec. Function-calling agents use this automatically.

Pricing

## Pay only for _what you use._

No subscriptions. No minimums. USDC on Base settles instantly via x402. The /probe endpoint is always free.

Static scraping

$0.005

per URL · Cheerio + Readability

-   No API keys or accounts
-   Pay per request via x402
-   USDC on Base mainnet
-   Article mode — strips ads & nav
-   Full mode — entire page
-   Free /probe endpoint

JS rendering

$0.015

per URL · Playwright + Chromium

-   Handles SPAs and dynamic pages
-   Bot-detection bypass
-   Same x402 payment flow
-   ~2–4s response time

**How x402 works:** The HTTP 402 status code requests micropayment. Your agent receives payment requirements, settles USDC on Base, and retries automatically.

[Learn about x402 →](https://x402.org/)

Why zlurp

## Unlike every other _scraping tool._

Firecrawl, Diffbot, and ScrapingBee all require accounts, API keys, or billing setup — none of which an AI agent can do autonomously.

Feature

zlurp

Firecrawl

Diffbot

ScrapingBee

No account required

✓

✗

✗

✗

No API key needed

✓

✗

✗

✗

x402 micropayments

✓

✗

✗

✗

Agent can start at 3am

✓

✗

✗

✗

Markdown output

✓

✓

✗

✗

JS rendering

✓

✓

✓

✓

OpenAPI spec

✓

✓

✓

✗

llms.txt

✓

✗

✗

✗