Read Webpage

Contents

Request
Response
Billing
Errors
Notes

Send a URL and get the page's main content: navigation, ads and footers removed, title and body kept, as Markdown or plain text. Useful for letting LLMs read web pages.

Request

POST https://platform.kuaisou.com/api/read-webpage
Authorization: Bearer YOUR_API_KEY
Content-Type: application/json

Parameters

Parameter	Type	Required	Description
url	string	Yes	The page URL, up to 1024 characters
format	string	No	`markdown` (default) or `json`; other values are treated as `markdown`

Example

curl -X POST https://platform.kuaisou.com/api/read-webpage \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"url":"https://www.example.com/article","format":"markdown"}'

Response

{
  "code": 200,
  "msg": "",
  "data": {
    "url": "https://www.example.com/article",
    "title": "Page title",
    "canonical": "https://www.example.com/article",
    "lang": "en",
    "status": 200,
    "format": "markdown",
    "quality": 0.92,
    "textChars": 3580,
    "warnings": [],
    "pageType": "article",
    "author": "Author",
    "siteName": "Site name",
    "date": "2025-06-01",
    "body": "# Page title\n\nBody text…"
  },
  "logId": "req_..."
}

Fields

Field	Description
url	The URL actually read (after redirects)
title	Page title
canonical	The canonical URL the page declares
lang	Page language
status	The HTTP status the site returned
format	The format you asked for
quality	Quality score of the extraction
textChars	Number of characters in the main content
warnings	Notes from the extraction (for example, the content may be incomplete)
pageType	Page type (article, list page, …)
author, siteName, date	Author, site name and publish date, when the page has them
body	The main content: Markdown when `format` is `markdown`, plain text when it is `json`
content	Only when `format` is `json`: the full extraction result

Billing

You are charged only when main content is actually extracted. If the page cannot be fetched, the site does not allow crawling, or no content is found, the API returns 422 and charges nothing. See the price table on the platform home page.

Errors

code	When
400	The body is not valid JSON, `url` is empty, or `url` is longer than 1024 characters
401	Invalid or disabled API key
402	Insufficient balance
422	Reading failed: `msg` says at which step and why; `data` has `step`, `status` (the site's status) and `url`
500	Reading the price, checking the balance or billing failed (not charged)
502	The reading service is temporarily unavailable (not charged)

See Errors.

Notes

We respect robots.txt: pages a site does not allow to be crawled cannot be read.
Some sites cannot be read for now; the API returns 422 with the reason.