Catalog

Web Scraping API

Turn any website into clean data: scrape a page to Markdown, map a site into its URLs, crawl whole sites into LLM-ready text, or capture real-browser screenshots. One key, one wallet, billed per request in real dollars.

Endpoints4Providers5From · scrape page$0.70/1k req · BisonFull breakdownRead docs
Uptime
94.03%
30d · 7,616 calls
Requests
7,642
30d · weekly, last 12 wks
Response
2.6s
median · 30d
POST

Scrape page

Most used
/v1/run/web.scrapeweb.scrape
Scrape any web page and get its main content back as clean Markdown plus title and metadata.
7,530 /mo
â–¼67.1%
93.95%
2561ms p50
4
from$0.70/1k req
Bison
POST

Map site

/v1/run/web.mapweb.map
Map an entire website into a clean list of its URLs, with titles and descriptions, in a single call.
102 /mo
â–²37%
99.02%
1469ms p50
1
price$1.00/1k req
Salamander
POST

Screenshot page

/v1/run/web.screenshotweb.screenshot
Capture a real-browser screenshot of any web page URL and get back a link to the rendered image.
9 /mo
â–²800%
100.00%
10021ms p50
1
price$1.74/1k req
Octopus
POST

Crawl website

/v1/run/web.crawlweb.crawl
Crawl a website and get clean text content from up to 10 pages in one normalized response - ideal for feeding sites into LLMs and search indexes.
1 /mo
â–²0%
100.00%
5283ms p50
1
price$3.15/1k resultsup to $0.0347 / request
Octopus

About the Web Scraping API

Every provider prices each endpoint independently, so the table shows the cheapest per endpoint. Open one to compare all providers.

Measured performance

What the Web Scraping API actually does

Every call through AnyAPI over the last 30 days, counted here rather than promised.

Latency percentiles measure time attributable to AnyAPI. Where an endpoint takes a caller-requested wait (for example web.scrape's waitFor), that wait is excluded, so your own wall-clock time is longer by whatever delay you ask for. Requests that use no such parameter are reported exactly as measured.

94.0%
of calls succeeded (7,616 calls)
Succeeded on first lane 88.7%Recovered by failover 5.3%Failed 6.0%
Response time estimate (7,161 calls across endpoints)
Weighted p50 estimate
2.6s
Request-weighted average of endpoint p50 values, not a combined request percentile
Weighted p99 estimate
19.9s
Request-weighted average of endpoint p99 values, not a combined request percentile

FAQ

About the Web Scraping API

The AnyAPI Web Scraping API returns Web Scraping data as normalized JSON across 4 endpoints: scrape page, map site, crawl website, and screenshot page. You call one endpoint per job with a single key, and AnyAPI routes each request across 5 sources, falling back automatically when one fails. Pricing is in US dollars, from $0.70 per 1,000 requests on the cheapest endpoint, with no subscription and no monthly minimum. Over the last 30 days, 94.0% of Web Scraping calls through AnyAPI succeeded, with a request-weighted median response time of 2.6 seconds across 7,616 measured calls.

Pricing is in US dollars, from $0.70 per 1,000 requests on the cheapest endpoint, with no subscription and no monthly minimum. You fund one USD wallet, each call draws it down, and a failed request costs $0.

4 endpoints: scrape page, map site, crawl website, and screenshot page. Every one takes the same authentication and returns the same normalized response shape, so reaching a second Web Scraping endpoint is a change of URL and nothing else.

Over the last 30 days, 94.0% of Web Scraping calls through AnyAPI succeeded, with a request-weighted median response time of 2.6 seconds across 7,616 measured calls. These are AnyAPI's own measurements of traffic through the gateway, recomputed continuously, not a published service-level target.