Fetch URL as Markdown

TypeScriptTools

POST api.upapi.io/fetch-markdown.post

Convert any web page into clean, LLM-ready Markdown. Fetches over plain HTTP with browser headers (retrying once through a residential or datacenter proxy when the origin serves a bot wall), removes boilerplate with Mozilla Readability, and converts with GitHub-flavored Markdown tables, lists and code blocks. Article mode is verified against a whole-page conversion and downgrades itself when Readability strips too much, so you are never handed a gutted page. Returns the markdown plus title, byline, language, excerpt, the extracted link list, the final URL after redirects, and which transport served the body. No browser is used, so a client-rendered page is reported as an error rather than as empty content.

markdownhtmlreadabilityllmragscrapercontentconverter

AuthAPI key

Cost3 units

Timeout25s

Cache300s

Request schema

Field	Type	Description
`url`req	string	Absolute http(s) URL of the page to convert (e.g. https://en.wikipedia.org/wiki/Markdown). Private, loopback and link-local destinations are refused.
`mode`	string	article = Mozilla Readability main-content extraction, verified against the whole page and auto-downgraded when it strips too much; full = the whole body with nav/footer/aside/form removed (e.g. article).· default: `"article"`
`maxChars`	integer	Maximum characters of markdown to return (e.g. 100000). Output is cut at the nearest paragraph/line boundary and `truncated` is set.· default: `100000`
`timeoutMs`	integer	Total network budget in milliseconds across the direct attempt and the one proxied retry (e.g. 20000). Capped below the 28s API-gateway edge timeout.· default: `20000`
`includeLinks`	boolean	Keep inline [text](url) links in the markdown and return the extracted link list (e.g. true). false unwraps every anchor to plain text.· default: `true`
`includeImages`	boolean	Keep ![alt](url) image references in the markdown (e.g. false). Images are dropped by default because they cost tokens and rarely help an LLM.· default: `false`

Response schema

Field	Type	Description
`lang`req	stringnull	Language tag from <html lang> (e.g. "en"); null if absent.
`links`	array	Absolute http(s) links found in the extracted content, capped at 500. Present only when includeLinks is true; reflects the extracted document, which may be longer than a truncated markdown body.
`title`req	string	Page/article title, or an empty string if none was found.
`byline`req	stringnull	Author line, when Readability or a meta tag exposes one; null otherwise.
`excerpt`req	stringnull	Short summary from Readability or the meta description; null if absent.
`finalUrl`req	string	URL actually fetched, after following redirects.
`markdown`req	string	The converted document. Prefixed with the page title as an H1 when the extracted content does not already start with a heading.
`charCount`req	integer	Length of the returned markdown.
`extractor`req	string	Which stage produced the markdown. In article mode this reads "full-dom" when Readability failed the content-density check.
`fetchedAt`req	string	ISO-8601 timestamp of the fetch.
`truncated`req	boolean	True when the markdown was cut to fit maxChars.
`httpStatus`req	integer	HTTP status of the response that produced the markdown.
`fetchedWith`req	string	Transport that produced the returned body: direct = straight from the worker; datacenter/residential = the proxy-pool tier used for the retry.

Try it

urlrequired

Absolute http(s) URL of the page to convert (e.g. https://en.wikipedia.org/wiki/Markdown). Private, loopback and link-local destinations are refused.

modeoptional· default: article

article = Mozilla Readability main-content extraction, verified against the whole page and auto-downgraded when it strips too much; full = the whole body with nav/footer/aside/form removed (e.g. article).

maxCharsoptional· default: 100000

Maximum characters of markdown to return (e.g. 100000). Output is cut at the nearest paragraph/line boundary and truncated is set.

timeoutMsoptional· default: 20000

Total network budget in milliseconds across the direct attempt and the one proxied retry (e.g. 20000). Capped below the 28s API-gateway edge timeout.

includeLinksoptional· default: true

Keep inline [text](url) links in the markdown and return the extracted link list (e.g. true). false unwraps every anchor to plain text.

includeImagesoptional· default: false

Keep ![alt](url) image references in the markdown (e.g. false). Images are dropped by default because they cost tokens and rarely help an LLM.

Run this endpoint live

Jump in as a guest — no signup — and run fetch-markdown.post with your own inputs. Create a free account anytime to get your own API key.

Code examples

curl -X POST https://api.upapi.io/fetch-markdown.post \
  -H 'Content-Type: application/json' \
  -H 'x-api-key: YOUR_API_KEY' \
  -d '{ "url": "https://en.wikipedia.org/wiki/Markdown", "mode": "article", "maxChars": 100000, "timeoutMs": 20000, "includeLinks": true, "includeImages": false }'

Developer tools

OpenAPI specImport the spec or collection into Postman, Insomnia, or any codegen tool.