NanoParse NanoParse

Parse the web.
$0.0025 at a time.

Pay-per-parse · USDC on Base · x402

NanoParse executes full JavaScript and converts any modern web app (SPA) into clean Markdown with rich metadata. No accounts or API keys—your AI agent simply calls our endpoint and pays $0.0025 USDC per parse from its own EVM wallet via the x402 protocol (Base network supported). Every IP gets 50 free parses to test first.

Get Started Try It Free
$0.0025
Per parse
x402
Payment
50
Free parses
Full JS
Edge Rendering
How it works

URL in. Rich data out.

Your agent sends a URL. We render it in a real browser, run intelligent content scoring to isolate the article, extract metadata, and return clean Markdown with author, dates, language, and structured data. Quarter cent per page.

01

Agent sends a URL

Your agent calls POST /fetch with any web URL. First 50 parses free per IP. Proxy rotation and abuse heuristics are actively monitored — sustained evasion triggers automatic blocking. No account, no API key, no human required.

02

402 — Payment Required

The API returns x402 payment terms. Your agent's wallet signs a $0.0025 USDC micropayment automatically.

03

Browser renders the page

Premium feature. Headless Chromium on the edge. Don't let your agents get stuck on empty React apps or infinite scroll. NanoParse fully executes JavaScript, waits for the network to idle, and perfectly renders complex SPAs before converting the DOM to Markdown.

04

Content scoring isolates the article

Smart block-level scoring detects navigation, footers, ads, and cookie banners — stripping them while keeping the article intact. Uniquely handles Obsidian-style callouts, math, and footnotes.

05

Metadata extracted

OpenGraph, Twitter Cards, schema.org JSON-LD, favicon, language, word count — everything scraped from the raw HTML and attached to the response.

06

Markdown delivered

Clean, structured Markdown returned with full metadata payload. Ready for your agent to consume, store, or analyze.

terminal
# Any agent can parse a page
$ curl -X POST https://nanoparse.app/fetch
-H "Content-Type: application/json"
-d '{"url":"https://example.com/article"}'

→ 402 Payment Required — $0.0025 USDC on Base
→ Agent signs payment with x402 wallet

$ curl -X POST https://nanoparse.app/fetch
-H "Content-Type: application/json"
-H "Payment-Signature: <signed-payload>"
-d '{"url":"https://example.com/article"}'

# Article Title
Full article content in clean Markdown.
No ads. No nav bars. No cookie popups.

$
Try It

Parse a page. Free.

3 free demo parses (no wallet needed). Then 50 free via the /fetch API. Turnstile protected — humans only.

Pricing

$0.0025 per page.

No subscriptions. No commitments. Pay exactly what you use. First 50 parses free.

Pay Per Parse
$0.0025 USD
Per page, paid in USDC via x402
MCP Server

One URL. Zero setup.

NanoParse runs an MCP server at nanoparse.app/mcp. Add it to your MCP client config and any AI agent discovers nanoparse_fetch as a native tool — with optional debug mode for inspecting extraction.

MCP config
// Add to your MCP client config
{
"mcpServers": {
"nanoparse": {
"url": "https://nanoparse.app/mcp"
}
}
}

// Agent auto-discovers: nanoparse_fetch(url)
// Optional: nanoparse_fetch(url, debug: true)
// → x402 payment: $0.0025 USDC
// → Markdown + metadata delivered

How MCP works

Add the NanoParse MCP server URL to your config. Your agent auto-discovers the nanoparse_fetch tool and calls it whenever it needs web content. Pass debug: true for detailed extraction diagnostics. The agent handles the x402 micropayment through its wallet — $0.0025 per page in USDC on Base.

API Reference

One endpoint. Rich metadata.

Works with any HTTP client. No SDK required. Just POST a URL and get back clean Markdown with author, dates, language, schema.org data, and more.

POST /fetch

Send a URL, get Markdown + metadata. First call returns 402 with payment instructions. Pay via x402, retry with signature, get full structured content. Pass "debug": true for extraction diagnostics.

https://nanoparse.app/fetch
request / response
POST /fetch

# Request body
{
"url": "https://example.com/article",
"debug": true // optional
}

# Response (200 — Paid)
{
"success": true,
"url": "...",
"markdown": "# Article Title Content...",
"metadata": {
"title": "Article Title",
"author": "Jane Chen",
"description": "...",
"published": "2025-01-01",
"image": "https://...",
"favicon": "https://...",
"site": "Example",
"domain": "example.com",
"wordCount": 1245,
"language": "en"
},
"schemaOrgData": { "@type": "Article", ... },
"freeCallsRemaining": 49,
"debug": { // when debug: true
"contentSelector": "auto-detected",
"metadataSource": "opengraph + schema.org + meta tags",
"cleaning": { "blocksRemoved": 2, "blocksKept": 8 },
"timing": { "metadataFetchMs": 180, "browserRenderMs": 2400 }
}
}

# Response (402 — Payment Required)
HTTP 402 Payment Required
Payment-Required: <base64-payment-instructions>
X-Free-Calls-Remaining: 0
FAQ

Questions agents would ask.

If they could. Since they can't, here's what their humans want to know.

Does this work on React, Vue, and heavy JavaScript sites?
Yes. Standard scrapers fail on modern Single Page Applications because they only fetch the raw HTML — you get back an empty <div id="root"></div>. NanoParse spins up a headless Chromium instance at the edge, fully executes all JavaScript, and waits for dynamic content to load before running our content isolation algorithms.
What is x402?
An open payment protocol backed by Coinbase and Cloudflare. It uses HTTP 402 — reserved since 1992 but never used — to let machines pay each other. An agent gets a 402 with payment terms, signs a USDC transaction, and retries.
Do I need a crypto wallet?
Yes. NanoParse uses the x402 protocol which requires USDC on the Base network (Ethereum L2). Here is what you need:

For testing: No wallet needed. Every IP gets 50 free parses.

For production — the dead-simple setup:
1. Install Coinbase Wallet (browser extension or mobile app).
2. Buy $5 of USDC — it arrives on Base network automatically.
3. Export your private key. Your agent script reads it from an env var:
export AGENT_WALLET_KEY="your-private-key-here"
4. That is it. Your agent is funded. $5 = 2,000 parses.

For AI agents: x402 handles wallet interactions automatically. Frameworks like x402 and agent SDKs sign transactions without human input. Your agent receives a 402, pays $0.0025, and retries in milliseconds.

Web2 developers: If you have never touched crypto before — you only need Coinbase Wallet, $5 of USDC, and one environment variable. No blockchain knowledge required.
What pages work?
Any public URL. JavaScript SPAs, docs, articles — full DOM rendering. Login-walled sites and sites blocking Cloudflare's IP range won't work.
How is this different from Firecrawl?
Firecrawl sells to humans: accounts, plans ($16–$599/mo), API keys. NanoParse is agent-native: no account, no plan, no key. Agent discovers endpoint, pays per-call, moves on.
Do you respect robots.txt?
Yes. We check robots.txt before every parse. Publishers: dmca@nanoparse.app for immediate blocklisting.
Works with Claude Code / Cursor?
Yes. MCP server at nanoparse.app/mcp. Add to config, agent auto-discovers the tool.
Why the atom logo?
Parsing = breaking complex structures into fundamental elements. The atom represents nano-scale precision. The cube represents structured data. Together: parsing the web at atomic scale.
Terms of Service

1. Service Description. NanoParse is a URL-to-Markdown conversion API. You submit a URL and we return the rendered page content as Markdown. The service is intended for use by AI agents and automated systems via the x402 micropayment protocol. NanoParse acts as a transient network conduit; we process requests dynamically, and any temporary caching of Markdown outputs is strictly for short-term performance and latency optimization.

2. Payment Terms. Usage is billed per request via the x402 protocol in USDC on the Base network. The current price is listed on our pricing page and may be updated with notice. You are responsible for all charges associated with your wallet address. All payments are final and non-refundable. NanoParse does not store, hold, or custody any cryptocurrency — payments are direct wallet-to-wallet. NanoParse is not responsible for failed, delayed, or misrouted transactions caused by underlying blockchain network conditions, smart contract glitches, or wallet provider errors on the Base network.

3. Acceptable Use. You agree not to use NanoParse to: (a) Violate any applicable law or third-party rights, including copyright; (b) Fetch content from sites that explicitly block automated access via robots.txt or other technical measures; (c) Bypass authentication, paywalls, or access controls; (d) Distribute fetched content in a way that misrepresents its source or ownership; (e) Exceed reasonable usage limits, launch denial-of-service attacks, or attempt to overload NanoParse infrastructure. You are solely responsible for how you use content delivered through NanoParse. Compliance with copyright law, fair use, and platform terms falls on you as the user of the API.

4. Blocklists. NanoParse maintains both a built-in blocklist of major paywalled news domains and a dynamic blocklist for any site that requests removal. These lists are enforced at the API level and cannot be overridden. Publishers: dmca@nanoparse.app.

5. Copyright & DMCA.

5.1 Notice. If you believe content accessible through NanoParse infringes your copyright, send a written notice to dmca@nanoparse.app containing: (i) your signature; (ii) identification of the copyrighted work; (iii) identification of the infringing material (the URL); (iv) your contact information; (v) a good faith statement that the use is not authorized; (vi) a statement under penalty of perjury that the information is accurate and that you are authorized to act on the copyright owner's behalf.

5.2 Takedown. Upon receipt of a valid notice, we will expeditiously remove or disable access to the identified material and notify the user who posted it.

5.3 Counter-Notice. If you believe material was removed due to mistake or misidentification, you may submit a counter-notice containing: (i) your signature; (ii) identification of the removed material and its location before removal; (iii) a statement under penalty of perjury of good faith belief that the removal was a mistake; (iv) your name, address, and phone number, and consent to federal court jurisdiction. We will forward the counter-notice to the original claimant. Unless the claimant informs us of a court action within 10-14 business days, we will restore the material.

5.4 DMCA Agent. NanoParse operates as a designated service provider under the Digital Millennium Copyright Act. Designated Agent Email: dmca@nanoparse.app. Registration: DMCA-1076317 (U.S. Copyright Office).

6. No Warranty. The service is provided "as is" and "as available" without warranty of any kind, express or implied. We do not guarantee that every URL will render successfully, that content will be complete or accurate, or that the service will be uninterrupted or error-free.

7. Limitation of Liability. To the maximum extent permitted by law, NanoParse and its operators shall not be liable for any indirect, incidental, special, consequential, or punitive damages arising from your use of the service. Our total liability for any claim shall not exceed the amount paid by you to NanoParse in the 12 months preceding the claim.

8. Indemnification. You agree to indemnify and hold NanoParse harmless from any claims, damages, or expenses arising from your use of the service or violation of these terms.

9. No Accounts & Privacy. NanoParse does not create accounts, store credentials, or collect personal information. The service is fully anonymous. No data about individual users is retained beyond aggregate usage statistics stored by month.

10. Changes. These terms may be updated at any time. We will indicate the date of the last update at the top or bottom of this document. Continued use after changes constitutes acceptance of the revised terms.

11. Governing Law & Dispute Resolution. These terms are governed by the laws of the State of Ohio, United States, without regard to conflict of law principles. Any dispute, claim, or controversy arising out of or relating to these terms shall be determined by mandatory binding individual arbitration in Ohio, rather than in court. You and NanoParse mutually waive any right to participate in a class action lawsuit or class-wide arbitration.

12. Contact. General: hello@nanoparse.app. DMCA: dmca@nanoparse.app.

Last updated: July 2026.

Privacy Policy

1. Information We Collect. IP Addresses — temporarily logged for rate limiting and abuse prevention, retained max 24 hours. Request URLs — processed in memory, not logged or stored.

2. Information We Do NOT Collect. No storage of extracted page content. No cookies, trackers, or analytics. No personal info unless you voluntarily contact us. No selling or sharing of data.

3. Email Communications. Emails to hello@nanoparse.app or dmca@nanoparse.app are used solely to respond. No mailing lists, no sharing. Delivery via Mailtrap.

4. Data Retention. IP logs deleted after 24 hours. Email retained only as long as necessary.

5. Third-Party Services. Cloudflare (CDN/DDoS) and Mailtrap (email delivery) may process limited data per their respective privacy policies.

6. Children's Privacy. Not directed at children under 13.

7. Contact. hello@nanoparse.app