Context for AI

Read a page

Give your model live web context. Hand the API a URL and get back clean, LLM-ready markdown. Hardened sites that block bots are handled automatically, and you are billed whenever a fetch is dispatched, whether or not it returns usable content.

Scrape a URL

Fetches a single page, strips navigation and boilerplate, and returns the main content as markdown, HTML, or plain text.

POST/v1/scrape

Parameters

urlstringrequired

The absolute http(s) URL to fetch. Private, loopback, and internal addresses are rejected.

formatstring

Output format: markdown (default), html, or text.

stealthboolean | "auto"

Controls the anti-bot rail. false (default) uses the fast standard rail only. "auto" starts standard and automatically escalates to the stealth rail if the site blocks the request or returns no content. true goes straight to stealth.

timeout_msinteger

Standard-rail timeout in milliseconds (1000–60000, default 30000). Stealth unlocking is allowed longer since it renders the full page.

Request

cURL

curl https://api.apexapi.dev/v1/scrape \
  -H "Authorization: Bearer ak-your-api-key" \
  -H "Content-Type: application/json" \
  -d '{
    "url": "https://example.com/pricing",
    "format": "markdown",
    "stealth": "auto"
  }'

Response

200 OK

{
  "url": "https://example.com/pricing",
  "final_url": "https://example.com/pricing",
  "status": 200,
  "format": "markdown",
  "content": "# Pricing\n\nSimple pricing for every team...",
  "metadata": {
    "title": "Pricing - Example",
    "description": "Simple pricing for every team.",
    "language": "en"
  },
  "tier": "standard",
  "cost": 0.002
}

The tier field tells you which rail served the page and therefore what you were charged. The response also carries X-ApexApi-Cost and X-ApexApi-Credits-Remaining headers.

Pricing

Billed at the price of whichever rail actually served (or attempted) the fetch. Only requests rejected before any fetch is dispatched, an invalid or disallowed URL, are free.

Tier	When	Price / page
Standard	Plain fetch, most sites	$0.002
Stealth	Anti-bot / JS render, hardened sites	$0.008

With stealth: "auto" you pay the standard price whenever the fast rail succeeds, and only pay the stealth price on pages that actually required unlocking.

Billing on failures

A request is billed whenever the fetch is dispatched to a rail, even if that fetch does not succeed. That includes a page that blocks the request, a target that is unreachable, and a page that returns HTTP 200 with no readable content to extract. Each of these reached the vendor and is billed at that rail's price.

Requests rejected before dispatch, such as an invalid, private, or disallowed URL, never reach a rail and are not billed.

Notes & limits

• Requests to private, loopback, or internal network addresses are rejected with a 400.
• The API respects each site's robots.txt by default; disallowed URLs return a 403.
• Use of this endpoint is subject to our Acceptable Use Policy. You are responsible for the URLs you submit and how you use the results.
• Need a whole site instead of one page? Use /v1/crawl to crawl a domain into markdown asynchronously.