Developing
The documentation does not give a user agent, IP ranges or robots.txt behaviour for scrape and crawl modes.
In 60 seconds
- Three modes: Scrape fetches one page, Map lists a site's URLs from links and sitemaps, Crawl follows links across up to 100 pages and three links deep.
- Crawl costs one credit per successfully extracted page; the free plan has 250 credits a month.
- It does not run JavaScript, so client-rendered content may be missing.
- Google sued SerpApi under the DMCA in December 2025; the original complaint was dismissed in July 2026 with leave to amend, and an amended complaint is pending.
Impact Layer
- What changed
- A SERP data company now sells general page and site extraction through the same endpoint.
- Why it matters
- Site owners may see new automated traffic, and AI tool builders get another route to web content.
- Who is affected
- Site owners, SEO and engineering teams and AI developers who need web data.
- Martech stack area
- SEO & Search, Data
- Stack impact
- Check your logs for SerpApi traffic and decide whether to block it, since its documentation does not state robots.txt handling for scrape and crawl.
- Business impact
- Low Based on PPC Land's reading of SerpApi's documentation. Neither Google's nor Reddit's claims directly concern the Extract API, and the docs page shows an unexplained "API uptime" of 82.823%.
- Related vendors
- What to watch
- The court ruling on Google's amended complaint and SerpApi's disclosure of its user agent.
SerpApi, an Austin company, released its Extract API to all accounts, including the free plan, on October 13, 2026, according to PPC Land.
What it does
- Returns pages as JSON, Markdown or HTML up to 5 MB, through the same endpoint as its search products.
- Scrape fetches one page; Map lists a site’s URLs from links and sitemaps; Crawl follows links across up to 100 pages, three links deep.
- Crawl costs one credit per successfully extracted page; failed pages and cached results are not charged. The free plan has 250 credits a month.
- It does not run JavaScript, so client-rendered content may be missing.
Not stated in the documentation
The user agent string or IP ranges, whether scrape or crawl modes obey robots.txt (only map mode reads it, for sitemap locations), crawl-delay handling, and per-domain rate limits.
Legal context
Google sued SerpApi in December 2025 under the DMCA. The court dismissed the original complaint in July 2026 with leave to amend; Google filed an amended complaint in August and SerpApi’s motion to dismiss remains pending. Reddit and SearchApi have separate suits. None of the claims directly concerns the Extract API.
Our read
If you do not want to be extracted, you need to know the user agent, and the docs do not give it. Watch your logs.
Sources
- PPC Land: SerpApi now scrapes any public web page, crawling up to 100 per request (opens in a new tab)
Source for the product, pricing, limits, legal context and gaps.