FetchDelta Quickstart

Fetch public webpages as clean markdown or diff them against the last snapshot. Paid endpoints use x402 on eip155:8453 with USDC settlement, so an agent can discover the service publicly and only pay when it actually needs fresh content.

Manifest JSON OpenAPI JSON llms.txt Health

Pricing

$0.002 /api/resolve per request, about 500 calls per $1

$0.004 /api/diff per request, about 250 calls per $1

Agent Buyer Flow

1. Inspect the endpoint Hit a paid URL and read the 402 Payment Required response headers.

2. Retry with an x402-capable client Use a Base wallet and pay exactly for the request you want.

3. Consume normalized JSON The unlocked response contains clean markdown or a structured diff, not raw page HTML.

Inspect the Paywall

curl -i 'https://fetchdelta.com/api/resolve?url=https%3A%2F%2Fexample.com%2Fpost'

Use the Included CLI

API_BASE_URL=https://fetchdelta.com node scripts/agent-client.js inspect resolve https://example.com

API_BASE_URL=https://fetchdelta.com EVM_PRIVATE_KEY=0x... \
node scripts/agent-client.js pay resolve https://example.com

The repository already includes scripts/agent-client.js, so buyers can inspect or pay with one command instead of wiring a client from scratch.

Paid Endpoints

/api/resolve at $0.002 per request
/api/diff at $0.004 per request

https://fetchdelta.com/api/resolve?url=https%3A%2F%2Fexample.com%2Fpost

https://fetchdelta.com/api/diff?url=https%3A%2F%2Fexample.com%2Fpost

Node Buyer Example

import { wrapFetchWithPaymentFromConfig } from "@x402/fetch";
import { ExactEvmScheme } from "@x402/evm";
import { privateKeyToAccount } from "viem/accounts";

const account = privateKeyToAccount(process.env.EVM_PRIVATE_KEY);
const paidFetch = wrapFetchWithPaymentFromConfig(fetch, {
  schemes: [{ network: "eip155:*", client: new ExactEvmScheme(account) }],
});

const response = await paidFetch("https://fetchdelta.com/api/resolve?url=https%3A%2F%2Fexample.com%2Fpost");
const data = await response.json();
console.log(data);

When to Use Which Endpoint

Use /api/resolve when an agent needs the current page body, clean markdown, title, headings, and metadata.

Use /api/diff when an agent already knows the page and only wants to know whether it changed since the last snapshot.