Scrape Response Fields

/v2/scrape returns a JSON payload:

{ "success": true, "data": { "markdown": "...", "metadata": { "title": "...", "statusCode": 200 } } }

Top-level

Field	Type	Notes
success	boolean	Whether the request succeeded
data	object	The result object (see below)

data (common)

Which keys appear depends on formats. See Output Formats.

Field	Type	Notes
markdown	string	Clean Markdown
html	string	Processed HTML
rawHtml	string	Raw HTML
links	string[]	Links found on the page
images	string[]	Image URLs found on the page
summary	string	LLM-generated summary
screenshot	string	Screenshot artifact (often a signed URL with expiry)
audio	string	Audio artifact (often a signed URL with expiry)
json	object	Structured extraction result ({ type: "json", ... })
metadata	object	Page metadata. See Metadata Fields.
warning	string	Warning message (e.g., truncation)

data.actions (when actions are used)

Field	Type	Notes
actions.screenshots	string[]	Screenshots produced by actions
actions.scrapes	array	HTML snapshots captured during actions ({ url, html })
actions.javascriptReturns	array	JS return values ({ type, value })
actions.pdfs	string[]	PDFs generated by actions

data.changeTracking (when changeTracking format is used)

Field	Type	Notes
changeTracking.previousScrapeAt	string (datetime)	Previous scrape time
changeTracking.changeStatus	string	Change status (e.g. new)
changeTracking.visibility	string	Visibility
changeTracking.diff	string	Diff (mode-dependent)
changeTracking.json	object	Structured diff (mode-dependent)

data.branding (when branding format is used)

branding is a large design-system profile object (colors, fonts, typography, spacing, components, etc.). Enable it only when you need brand/design extraction.