Web Data
The web, ready
for your workflow.
Collect pages, extract useful information and keep track of what changes.
From URL to useful data
Collect. Track. Inspect.
Scrape public HTTP pages, or request JavaScript rendering with bounded browser actions.
Poll asynchronous jobs and review per-URL outcomes from batch, crawl and map requests.
Read capture metadata and retrieve stored artifacts with integrity checks.
Clear collection boundaries
Know what each run covers.
Batch requests accept up to 100 URLs. Crawl and map requests stay on the same origin and support up to 500 pages at depth 3. Results preserve partial and failed page outcomes so callers can decide what to retry.
Read the API contract ↗Current capture output
Read page content, capture metadata, timestamps and available artifacts. Browser rendering and screenshots are available through bounded options.
Limits are part of the contract
These are per-request bounds. Workspace, project and scoped-key limits apply alongside plan allowances. A collection limit is not a freshness guarantee.
Saved evidence workflows
Extract, monitor and review.
- Typed CSS field extraction and reusable project recipes
- PDF text and OCR in seven supported languages with per-page methods
- Page monitoring, content diffs and recurring collections
- Evidence-backed page and automatic bounded crawl audits
- Imported Google SERPs (US, English, desktop) and domain comparisons
Audits retain incomplete and unavailable evidence. Imported search data is unverified; observed gaps do not establish global keyword coverage.
Read the evidence contractsAvailability
Collect pages with scrape, batch, crawl and map. Extract fields, monitor changes and review saved evidence in your workspace.
Connect Search Console for properties you can access. Native search is in preview; verified live rank tracking and other Google datasets have separate availability requirements.
Data sources and limits →Plans and usage
Choose the allowance
that fits your workflow.
Start with the Free allowance and explore larger monthly plans as your workflow grows.
See plans →What counts as a unit?
A page reserves one unit; native search reserves two. Batches reserve one per URL; crawls and maps reserve their requested maximum pages. Accepted failures and cancellations do not refund units. History reads do not consume collection units.
Free resets each UTC calendar month; paid allowances follow your subscription period. No automatic overage, annual billing or rollover. Provider and model requests use separate operator allowances.