Scrape

Read one page. Get clean Markdown.

Scrape reads one web page and keeps its text as Markdown, with the capture it came from.

Free. No sign-up. No card.

Scrapenorthwind.example/blog/launch-week
  • # Launch week
  •  
  • Five releases in five days. Here is what shipped,
  • and what is next.
  •  
  • - Monday: faster sync for large teams
  • - Tuesday: shared views
  • - Wednesday: audit history
Read 1 page · Markdown, with its source

How it works

From address to Markdown.

What Kaiju does with the address you send.

01

Opens the page

Over plain HTTP first. A browser opens only when the page needs one to show its text.

02

Keeps the content

Menus, headers and footers stay out. Or keep the whole page, or only the parts you name.

03

Returns Markdown and its source

Headings, lists, links and tables stay. Each result names the final address and a hash of the copy.

What you get back

One result: the page and its source.

Each step lights the part of the example that answers it. Pick one, or let it play.

Step 1 of 4
GET/v1/operations/{id}/results/{resultId}
Example
{  "id": "3f1a9c0e…",  "operationId": "0b7f3c1e…",  "kind": "scrape",  "sequence": 1,  "status": "ok",  "url": "https://example.com/",  "contractVersion": "1",  "bytes": 6120,  "sha256": "<64 hex characters>",  "markdown": "# Example Domain\n\nThis domain is for use in documentation examples.",  "diagnostics": [],  "source": {    "finalUrl": "https://example.com/",    "representation": "decoded_html",    "encoding": "utf-8",    "bytes": 52110,    "captureId": "<capture id>",    "interpreterId": "<interpreter id>",    "sha256": "<64 hex characters>"  }}

Build a request

Every option, with its default.

Change an option and the request changes with it. Run it here, free, then keep going in the playground.

The example site is made up. Any public address works in the playground.

Main options
KeepDefault: Main content
OutputDefault: Markdown
More options6
Content
Keep only these partsUp to 100 CSS selectors
Leave out these partsUp to 100 CSS selectors
Include extracted linksDefault: off
Page accessShared by Scrape, Map and Crawl.
Open pages withDefault: Automatic
Wait after page loadDefault: 0 ms · 0 to 30,000
ms
Time limitDefault: 30 s · 5 to 120
s
Reuse a saved copy up toPlannedNow: Fresh every run
Planned: 0 to 720 h
Mobile browserPlanned
Planned

Every option has a default Change one and the request updates. Focus an option to read what it does.

POST /v1/scrape
{
  "url": "https://northwind.example/blog/launch-week"
}

Free. No sign-up. Run it here, then keep going in the playground.

Which endpoint

Pick the endpoint for the job.

Start from what you have. Each endpoint does one job.

You know the page you need.

You send
One page address
You get
The page as Markdown, with its source
Limit
One page for each job.
You are on this page
Smallest request
POST /v1/scrape
{
  "url": "https://northwind.example/blog/launch-week"
}

You need the list of pages before you read them.

You send
A site, or a section of it
You get
A list of page addresses
Limit
Set by the service. Read GET /v1/capabilities.
See Map
Smallest request
POST /v1/map
{
  "url": "https://northwind.example",
  "maxUrls": 100
}

You need the text of many pages in one section.

You send
A start page and your limits
You get
One result for each page read
Limit
Set by the service. Read GET /v1/capabilities.
See Crawl
Smallest request
POST /v1/crawl
{
  "url": "https://northwind.example/docs/",
  "maxPages": 5,
  "maxDepth": 1
}

You need named values, such as prices, not whole pages.

You send
Up to four pages and the fields you name
You get
One record with those values
Limit
Up to 4 pages and 20 fields for each job.
See Extract
Smallest request
POST /v1/collect
// 503 capability_unavailable: not available yet

Cost and limits

Pay for the page Kaiju reads.

Draft rates. We confirm them before release.

1 credit per page

Each page

The draft rate for each page Kaiju reads. We confirm it before release.

1 page

Each job

One page for each job. The time limit covers the whole job, not each page.

0 credits

Not charged

Pages outside your scope: not fetched, so not charged. Pages past your limit: not read, not charged, and listed for you.

Work out a whole job in the pricing estimator.

Try Scrape

Read a sample page now.

The playground is free and needs no sign-up.

Privacy choices

Optional website tracking

These choices control optional browser tracking on Kaiju's public website. The app asks separately for identified product analytics. Turning Analytics off here also stops optional app analytics. These choices do not change information you intentionally submit through a contact, evaluation, or meeting form.

Essential

Website delivery, security, forms, and this saved preference.

Always on