Overview
Turn any public URL into clean, structured content over a single HTTP call.
The Scrape API fetches any public web page and returns its content as clean, normalized output. Navigation, ads, cookie banners, and boilerplate are stripped. What comes back is ready to feed into a language model, a vector store, a pipeline, or any system that works better with text than with raw HTML.
When to use it
- Feeding web content into an LLM context window
- Building a retrieval pipeline over live or crawled pages
- Extracting structured data from pages that do not expose an API
- Powering an AI agent that needs to read the web
What comes back
The API returns the page in the format you asked for. Request markdown and you get clean prose with headings, lists, and code blocks preserved. Request HTML and you get the stripped article content without the page chrome. Request JSON and you get a structured representation. Request binary and you get the raw bytes.
See Output formats for the full breakdown.
Paying for calls
Each call is billed per unit. The base price is $0.002 per page. Options like ISP proxying or JavaScript rendering multiply the unit count. Payment runs over MPP with your Tempo wallet. No account, no invoice. See How it works for the full payment flow.
Get started
Go to the Quickstart to make your first call.