Getting started

JsonLink API documentation

A simple REST API to get structured data from any web page. Every endpoint takes a URL and your API key and returns JSON.

Quickstart

Create a free account and copy your API key from the dashboard. The free plan includes 100 credits every month.
Call any endpoint with the url you want to process and your api_key:

Request

curl "https://jsonlink.io/api/extract?url=https://github.com&api_key=YOUR_API_KEY"

Read the JSON response:

Response

{
  "title": "GitHub · Change is constant. GitHub keeps you ahead. · GitHub",
  "description": "Join the world's most widely adopted, AI-powered developer platform where millions of developers, businesses, and the largest open source community build software that advances humanity.",
  "images": [
    "https://images.ctfassets.net/8aevphvgewt8/4pe4eOtUJ0ARpZRE4fNekf/f52b1f9c52f059a33170229883731ed0/GH-Homepage-Universe-img.png"
  ],
  "sitename": "GitHub",
  "favicon": "https://github.com/favicon.ico",
  "domain": "github.com",
  "url": "https://github.com",
  "duration": 1240
}

Endpoints

All endpoints live under https://jsonlink.io/api and share the same key and credits.

Endpoint	Returns	Credits
`GET /api/extract`	Title, description, images, favicon and site name of a URL.	1
`GET /api/screenshot`	Screenshot of a page, desktop, mobile or full page.	2
`POST /api/markdown`	Main content of a page as clean Markdown, for LLMs and RAG.	1
`GET /api/links`	Every link on a page, with text, rel and internal/external.	1
`GET /api/images`	Every image on a page, with format, dimensions and size.	1

Basics

Send the full page URL, including https://, and URL-encode it in query strings.
Responses are JSON. Extraction problems come back as an object with an error field — always check for it. See Errors.
Only successful requests use credits. Timeouts and pages that can't be loaded are free.
Pages are loaded in a real browser when needed, so JavaScript-heavy sites work. Pages behind a login are not supported.

Using JsonLink with AI

Connect the JsonLink MCP server to Claude, Cursor or VS Code and your agent can read pages, take screenshots and get link metadata on its own. The Markdown endpoint turns any page into clean text for LLMs and RAG pipelines, and a machine-readable summary of these docs is at /llms.txt.