ExtractSoon

Name the values. Get only those back.

Extract will return only the values you name, as one record. It is not available yet.

Extract is not available yet. The other three are free in the playground, with no sign-up.

Extractnorthwind.example/products/kettle
  • Price$89Accepted
  • StockIn stockAccepted
  • ShipsMar 3 or Mar 4Conflict
  • WarrantyNoneUnknown
An example of the record Extract will return

How it works

From a field name to a checked value.

How Extract will find each value.

01

You name the values

List the fields you want, such as a price, a date or a stock level.

02

Kaiju finds them

It reads up to four pages and looks for each value.

03

Each value shows its evidence

A value comes back with the text it came from, or as unknown. It is never guessed.

What you get back

A request that fails today.

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

Step 1 of 2
POST/v1/collect · 503
Example
{  "type": "urn:kaiju:crawl:capability_unavailable",  "title": "capability unavailable",  "status": 503,  "code": "capability_unavailable",  "correlation": "c2f1d7a06b9e"}

Build a request

Options open when it ships.

Extract does not take requests yet.

Extract is not available yet. Its fields and evidence options open here when it ships. Until then, Scrape, Map and Crawl run in the playground, free.

Try Scrape

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.
See Scrape
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.
You are on this page
Smallest request
POST /v1/collect
// 503 capability_unavailable: not available yet

Cost and limits

The planned rate.

Draft rates. We confirm them before release.

1 credit per page fetched

Each page

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

4 pages

Each job

Up to 4 pages and 20 fields 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.

Coming soon

Start with Scrape, Map and Crawl today.

Extract is not available yet. The other three run in the playground, free and with 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