Quickstart

Register for a free API key at captureweb.dev/signup. No credit card required. You get 100 credits to start.

curl -X POST https://api.captureweb.dev/v1/screenshot \
  -H "Authorization: Bearer pk_live_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{"url":"https://stripe.com","format":"png"}' \
  --output screenshot.png

const res = await fetch('https://api.captureweb.dev/v1/screenshot', {
  method: 'POST',
  headers: {
    'Authorization': 'Bearer pk_live_YOUR_KEY',
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({ url: 'https://stripe.com', format: 'png' }),
});
const png = await res.arrayBuffer();
// write to file or pass to your image CDN

import requests

res = requests.post(
    'https://api.captureweb.dev/v1/screenshot',
    headers={'Authorization': 'Bearer pk_live_YOUR_KEY'},
    json={'url': 'https://stripe.com', 'format': 'png'},
)
with open('screenshot.png', 'wb') as f:
    f.write(res.content)

req, _ := http.NewRequest("POST", "https://api.captureweb.dev/v1/screenshot",
    strings.NewReader(`{"url":"https://stripe.com","format":"png"}`))
req.Header.Set("Authorization", "Bearer pk_live_YOUR_KEY")
req.Header.Set("Content-Type", "application/json")
resp, _ := http.DefaultClient.Do(req)

Authentication

Pass your API key in the Authorization header on every request:

Authorization: Bearer pk_live_YOUR_KEY

Keys start with pk_live_. Get yours at /signup. Store it securely — it's shown once.

Credits

Credits are consumed per successful render:

Response headers on every request: x-credits-used, x-credits-remaining, x-cache (hit/miss), x-render-source, x-render-time.

Errors

/screenshot

/pdf

Accepts either url or html. For HTML input, the PDF is generated directly from the raw HTML — no URL needed.

/snapshot

Cheapest way to get all formats. One call returns screenshot, PDF, HTML, and Markdown together — faster and cheaper than calling each endpoint separately.

/scrape

/markdown

Strips navigation, ads, and boilerplate. LLM-ready output.

/content

/links

/json

/a11y

Extracts the full ARIA accessibility tree after JavaScript rendering. Useful for UI automation, accessibility audits, and giving AI agents a structured, semantic page representation.

Response: { url, tree, title, source, cached } where tree is the nested ARIA node object.

/brand

Extracts brand assets from any domain using a plain HTTP fetch — no browser required. Fast (~300ms) and reliable. Ideal for directories, CRMs, and any product that needs brand metadata at scale.

Response: { url, favicon, logo, og_image, og_title, og_description, colors[], fonts[], twitter_card, source, method, timing_ms, credits }

• colors — top brand colors as hex strings, sorted by frequency
• fonts — font families found in CSS and Google Fonts links
• logo — best candidate logo URL (apple-touch-icon → img[src*=logo] → OG image)
• favicon — resolved favicon URL

Async mode

Any endpoint accepts "async": true. The request returns immediately with a job ID. Poll GET /v1/jobs/:id or receive the result via webhook.

{
  "url": "https://example.com",
  "async": true,
  "webhook_url": "https://yoursite.com/hook"
}

Job status values: pending → running → completed | failed

/crawl

Render chain

Every request is automatically routed through up to 5 tiers. You never configure this — it's automatic.

The x-render-source header on every response tells you which tier was used.

Rate limits

429 responses include retry_after_seconds in the JSON body. Rate limits reset every minute on a sliding window.

Agents & MCP

CaptureWeb ships an MCP (Model Context Protocol) server that gives Claude, Cursor, Codex, and any MCP-compatible client the ability to see the web — screenshots, PDFs, clean Markdown, structured JSON, and accessibility trees from any URL.

Install via npx — no separate install step needed. The server connects over stdio and exposes all CaptureWeb endpoints as first-class MCP tools.

Claude Desktop

Open ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) and add:

{
  "mcpServers": {
    "captureweb": {
      "command": "npx",
      "args": ["-y", "@captureweb/mcp"],
      "env": {
        "CAPTUREWEB_API_KEY": "pk_live_YOUR_KEY"
      }
    }
  }
}

Restart Claude Desktop. CaptureWeb tools appear in the tool picker automatically.

Cursor

Open Cursor → Settings → MCP and add:

{
  "captureweb": {
    "command": "npx",
    "args": ["-y", "@captureweb/mcp"],
    "env": {
      "CAPTUREWEB_API_KEY": "pk_live_YOUR_KEY"
    }
  }
}

MCP tools

Every CaptureWeb endpoint is exposed as an MCP tool. Claude can call these directly when you ask it to capture or extract from a URL.

Example prompts once connected:

• "Take a screenshot of stripe.com/pricing and describe the layout"
• "Extract all product names and prices from amazon.com as JSON"
• "Get the markdown from the OpenAI blog and summarise the last 5 posts"
• "Check the accessibility tree of github.com — is the main nav keyboard-accessible?"
• "Take a full-page screenshot of linear.app in dark mode"

Skills

A Skill is a YAML file that gives any Claude Code project instant knowledge of the CaptureWeb API — no MCP server required. Drop it in your project and use it as a slash command.

Download captureweb-skill.yaml and save it to .claude/commands/captureweb.md in your project. Then in any Claude Code session, type /captureweb to activate it.

The skill includes the full endpoint reference, auth header format, and caching behaviour — so Claude knows exactly how to call the API without any extra context from you.

Agents.md

captureweb.dev/agents.md is a plain-text reference designed to be fed directly into any AI agent. It covers MCP setup, all 10 tools, REST examples in Python and TypeScript, and caching behaviour.

Paste this into your agent to let it handle setup automatically:

curl -s https://captureweb.dev/agents.md | claude -p "Follow these setup instructions"

Or open the file directly and copy it into any system prompt, CLAUDE.md, or agent context window.

Using the REST API in agent code

Prefer calling the API directly from your agent? Use the Authorization: Bearer header on any endpoint.

import os, requests

API = "https://api.captureweb.dev"
KEY = os.environ["CAPTUREWEB_API_KEY"]

# Extract Markdown for RAG
res = requests.post(f"{API}/v1/markdown",
    headers={"Authorization": f"Bearer {KEY}"},
    json={"url": "https://openai.com/blog"}
)
markdown = res.json()["markdown"]

# Screenshot
res = requests.post(f"{API}/v1/screenshot",
    headers={"Authorization": f"Bearer {KEY}"},
    json={"url": "https://stripe.com/pricing", "full_page": True}
)
open("screenshot.png", "wb").write(res.content)

const API = "https://api.captureweb.dev"
const KEY = process.env.CAPTUREWEB_API_KEY

// Accessibility tree for agent navigation
const res = await fetch(`${API}/v1/a11y`, {
  method: "POST",
  headers: { Authorization: `Bearer ${KEY}`, "Content-Type": "application/json" },
  body: JSON.stringify({ url: "https://github.com" }),
})
const { tree } = await res.json()

Cache is automatic — repeated calls to the same URL within 24 hours are free. Pass "fresh": true to force a new render.