BASECRAWL DOCS

Build a dependable pipeline from the open web.

Start with one page, scale to a whole site, and receive predictable markdown, JSON, or HTML that is ready for your application and your models.

RECOMMENDED START

Begin with Quickstart.

Make one authenticated request before you compare endpoints, configure a crawl budget, or choose an SDK.

Start Quickstart
  1. 01

    Create an API key

    Create one key to identify your project and authorize every server-side request.

  2. 02

    Send your first request

    Start with one URL and return normalized content in the format you need.

  3. 03

    Use the response

    Pass the canonical document into your product, pipeline, or model context.

API reference

Understand request conventions, output formats, pagination, and job states.

Explore the API

Crawl

Map a site, control depth, and receive normalized output page by page.

Crawl a site

Scrape

Extract focused content from one page with rendering controls and clean output.

Scrape a page

SDKs

Use typed JavaScript and Python clients with the same API semantics.

Install an SDK

PRODUCT FIGURE

One authenticated scrape before you expand scope.

Treat Quickstart as a product surface: headers, formats, and a clean document contract your pipeline can trust.

Flow diagram

Authenticate once, scrape one public URL, then decide whether you need /crawl depth and budgets.

docs · request path

Start here

/docs/quickstart

Curl, JavaScript, and Python samples with the same response contract.

Then expand

/crawl · /map · SDKs

Add depth budgets and typed clients after the first page succeeds.

First successful path

  1. 01

    Auth key

    bc_live_… header

  2. 02

    POST /v1/scrape

    one URL body

  3. 03

    Pick formats

    md · json · html

  4. 04

    LLM-ready doc

    stable contract

UI product frame

A single authenticated scrape — dark response pane nested inside a flat paper window frame.

terminal · first request

requestPOST · /v1/scrape
URLhttps://docs.northstack.dev/pricing
markdownjsonhtml
shellcurl · 200 OK
curl -X POST https://api.basecrawl.dev/v1/scrape \
  -H "Authorization: Bearer bc_live_demo" \
  -H "Content-Type: application/json" \
  -d '{
    "url": "https://docs.northstack.dev/pricing",
    "formats": ["markdown", "json"]
  }'
response.markdownclean · 3.8 kB
# Pricing

Growth ships 20k pages/mo.

- Clean RAG signal
- Source URL + status
- No leftover nav chrome

FIRST REQUEST

Start with a single URL, not a platform migration.

Keep your existing stack. Basecrawl returns stable documents through an API surface designed for production workflows.

Read quickstart