SerpApi is built around search engines, and it's good at them. Point it at Google, Bing, or Amazon and it hands back the results page as clean, structured JSON, with the CAPTCHAs solved and the proxies handled for you.
The question is what happens when the data you want isn't on a results page. Since October 2026, SerpApi has an Extract API that fetches a public page and parses it into structured JSON or Markdown. It doesn't run JavaScript, it follows robots.txt, and it returns SerpApi's standard fields. Anakin starts from the page instead: it renders JavaScript, keeps you logged in with Browser Sessions, clicks and scrolls before capture, and returns JSON in the schema you define. So here's the short answer. For ranked search results with precise location control, pick SerpApi. For the page behind the result, pick Anakin. The sections below cover scope, output, search, reach and cost, in that order.
What is each tool built for?
SerpApi is a search-results API. You pick an engine, pass a query, and get parsed results back. It covers more than 20 engines, including Google Search, Images, News, Maps, Shopping, and Scholar, plus Bing, DuckDuckGo, Baidu, Yandex, and verticals like Amazon, eBay, Walmart, and the Apple App Store. The design is engine-first: you choose an engine parameter, and even the new Extract API is just another engine (engine=extract) that takes a URL. Search pages that have a dedicated engine get redirected to it.
Anakin is a general web-data platform. The URL Scraper takes any URL and returns the page. Crawl walks a whole site, Map discovers its links, Browser Sessions hold an authenticated login, the Browser API drives a real browser, and a Search API covers the discovery half of the job. Point Anakin at a product page, a listing behind a login, or a JavaScript-heavy app, and it returns the content, or structured JSON shaped to your schema.
This is where the two split. A SERP request returns a list of result URLs, and those pages still have to be fetched and parsed. SerpApi's Extract API now covers the simple case: a public, server-rendered page, parsed into SerpApi's standard sections. A listing that only renders in the browser, a page that robots.txt disallows, or a supplier portal behind a login is outside what it does.
Anakin starts from the page. You give it a URL, then add useBrowser for JavaScript-heavy pages, a sessionId from Browser Sessions for pages behind a login, actions to click or scroll first, and an outputSchema for the exact fields you want back. Here it is pointed at SerpApi's own pricing page, asking for each plan's name, price, monthly searches and hourly throughput:
# 1. Submit the URL plus the fields you want back
curl -s -X POST https://api.anakin.io/v1/url-scraper \
-H "X-API-Key: $ANAKIN_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"url": "https://serpapi.com/pricing",
"country": "us",
"outputSchema": {
"type": "object",
"properties": {
"plans": {
"type": "array",
"items": {
"type": "object",
"properties": {
"name": {"type": "string"},
"price_per_month": {"type": "string"},
"searches_per_month": {"type": "string"},
"throughput_per_hour": {"type": "string"}
},
"required": ["name", "price_per_month", "searches_per_month", "throughput_per_hour"]
}
}
},
"required": ["plans"]
}
}'
# returns {"jobId": "...", "status": "pending"}
# 2. Poll with the jobId until "status" is "completed"
curl -s https://api.anakin.io/v1/url-scraper/$JOB_ID \
-H "X-API-Key: $ANAKIN_API_KEY"A run on 25 Oct 2026 returned all six plan cards on the page: Free, Starter, Developer, Production, Big Data and Enterprise. Here are the first three, trimmed:
{
"status": "completed",
"cached": false,
"generatedJson": {
"status": "success",
"data": {
"plans": [
{"name": "Starter", "price_per_month": "$25",
"searches_per_month": "1,000 searches per month",
"throughput_per_hour": "200 throughput per hour"},
{"name": "Developer", "price_per_month": "$75",
"searches_per_month": "5,000 searches per month",
"throughput_per_hour": "1,000 throughput per hour"},
{"name": "Production", "price_per_month": "$150",
"searches_per_month": "15,000 searches per month",
"throughput_per_hour": "3,000 throughput per hour"},
...
]
}
}
}The call returns a jobId. Poll GET /v1/url-scraper/{id} until the status reads completed. Because the request carried an outputSchema, the fields come back in generatedJson, shaped to that schema, on any page you choose, not a preset list. Schema extraction adds two credits to the one-credit scrape.
What do you get back?
Reaching the page is half of it. The other half is the shape of what comes back. SerpApi returns parsed SERP fields as JSON, and also offers raw HTML and a Markdown output tuned for LLMs. The Extract API applies the same idea to any page it fetches, returning fixed sections such as headings, tables, links and price offers.
Anakin returns HTML, Markdown, or schema-constrained JSON through the outputSchema parameter. The difference is whose structure you get. SerpApi returns its own fields, for search results and now for extracted pages. Anakin returns the fields you ask for. Website Monitoring builds on the same schema: a specific_data monitor diffs only the fields in your outputSchema, at 3 credits a check.
Search: a results page, or a researched answer?
Search is where the two products overlap most, and where the difference is easy to miss. SerpApi returns the search results page: titles, links, snippets, and the structured fields around them. That's the SERP, and for rank tracking it's exactly right.
Anakin covers both halves, under one API key. The Search API handles discovery: one synchronous call, three credits, and back come ranked results with URLs, titles, dates and text snippets. When you need the full page behind a result, hand its URL to the URL Scraper. When you need an answer rather than a reading list, Agentic Search does the reading for you. It refines the query, searches, scrapes the cited pages for their full content, and writes a cited report from 20+ sources, at 10 credits plus one per page it scrapes. AI Visibility takes the same idea to answer engines, sending one prompt to ChatGPT, Gemini, and Google AI Overview and returning every engine's answer side by side. None of it replaces a rank tracker. It's the discovery half of the job, done for an agent rather than a dashboard.
Where can each one reach?
Give SerpApi its due here. Its location control is genuinely strong: city-level targeting, GPS coordinates, encoded UULE values, and device selection make it precise for measuring how a SERP looks from a specific place. For rank tracking that granularity is what you're paying for.
Anakin's coverage is built for a different job. It reaches 207 countries and carries the anti-bot infrastructure to get through protections like Cloudflare, DataDome, and Akamai on pages that actively fight scrapers.
What does it cost to start and to scale?
Scope, output and reach decide which tool fits. Cost decides how it scales. SerpApi sells monthly subscriptions where one request equals one "search." Plans run from Free at 250 searches to Production at $150 for 15,000, and up into the enterprise tiers. Only successful searches count against your total, which is fair. The model rewards steady, predictable volume and is less kind to variable workloads: there's no pay-as-you-go option, unused searches don't roll over, and once a plan hits its cap you either wait for the next cycle, upgrade, or turn on Automatic Early Renewal, which starts a new paid cycle early. A workload that spikes and dips pays every month for headroom it never touches.
Anakin bills per credit. A scrape is one credit, JavaScript rendering is included rather than a surcharge, and you're charged only on success. Failed requests, timeouts and cache hits cost nothing. Plan credits refill each month, but top-up credits never expire, so a spiky workload can run a small plan and top up for the peaks. Funded startups can get up to 400,000 pages and 50% off for six months. The live per-plan numbers sit on the Anakin pricing page.
| SerpApi | Anakin | |
|---|---|---|
| Primary scope | Search engines, plus basic page extraction | Any URL on the web |
| What you target | An engine parameter | A URL |
| Billing | Monthly search subscription | Per credit, pay-as-you-go |
| Failed requests | Not counted | Free |
| Unused quota | Expires monthly | Plan credits refill monthly, top-ups never expire |
| Structured JSON | SerpApi's standard fields | Your own schema, any page |
| Search returns | Links and snippets | Snippets (Search API), or a cited report from full pages (Agentic Search) |
| JavaScript-rendered pages | Extract API doesn't run JavaScript | Headless browser (useBrowser) |
| Pages behind a login | Not documented | Browser Sessions |
| Anti-bot on arbitrary sites | Not documented for the Extract API | Beats Cloudflare, DataDome, Akamai |
| Compliance | SOC 2 Type II, ISO 27001 | SOC 2 Type II, ISO 27001:2022 |
Values are editorial assessments based on each vendor's public documentation as of October 2026, not independently benchmarked figures.
SerpApi gates throughput per plan, from 50 searches an hour on the free tier to 20,000 an hour on the $725 Searcher plan and hundreds of thousands on the largest Cloud plans. Anakin's Zero Touch mode takes the friction off the front of that curve: scrape a page or run a read-only Wire action with no key and no signup, metered per IP, then upgrade in place by adding an X-API-Key header when the project is ready. From there, concurrency scales with the plan, from 5 concurrent requests on the free Starter plan to 200 on Business.
Compliance and the legal question
Both companies hold SOC 2 Type II and ISO 27001, so the core certifications don't separate them. SerpApi's real edge is its U.S. Legal Shield: up to $2M of assumed liability for lawful search-engine scraping, included on Production plans and higher. If legal indemnity for SERP collection is the deciding factor for your team, that's a real SerpApi advantage worth naming.
The backdrop is an unsettled area of law. Google sued SerpApi under the DMCA, and those claims were dismissed in July 2026. It isn't over: Google filed a narrower amended complaint in August, and SerpApi has asked the court to dismiss that one too. SerpApi leans into being the vendor that carries that risk on your behalf, which is a coherent position for a SERP specialist.
Developer experience
Neither tool makes you fight the API. SerpApi offers a REST endpoint, SDKs across ten languages, and an MCP integration so an agent can call it as a tool. Anakin ships SDKs for the major languages (the docs list Python, Node.js, Go, Rust, PHP, Ruby, Elixir and Java), a CLI, and an MCP server exposing its full surface, so the URL Scraper, Search, and Wire calls are all available to a coding agent behind a single key.
When SerpApi is the right call
If the job is search-results data at volume, rank tracking, SERP monitoring, shopping or maps results, and you want city-level location control plus a legal shield, SerpApi is purpose-built for it. It's a specialist, and specialists earn their place.
Frequently Asked Questions
Can SerpApi scrape any URL?
Partly. Since October 2026 its Extract API fetches a public web page and returns structured JSON or Markdown. It doesn't run JavaScript, it follows robots.txt, and it returns SerpApi's standard fields rather than a schema you define.
How does SerpApi count searches against a plan?
Only successful searches count toward the monthly total. Plans run from Free at 250 searches a month to Production at $150 for 15,000, with larger tiers above that. Unused searches don't roll over, and there's no pay-as-you-go option.
How does Anakin bill for scraping?
Anakin bills per credit. A page scrape is one credit, JavaScript rendering isn't a surcharge, and you're charged only when a request succeeds, so failed requests and cache hits cost nothing. Top-ups are pay-as-you-go, and funded startups can get up to 400,000 pages and 50% off for six months.
Which SerpApi plans include the U.S. Legal Shield?
The U.S. Legal Shield is included on the Production plan and higher. It covers up to $2M of liability for the lawful scraping and parsing of search engine data.
Can Anakin return structured JSON from any page?
Yes. Send a URL to the URL Scraper and add an outputSchema to the request, and Anakin returns JSON shaped to that schema instead of raw content. It also returns HTML or Markdown if that's what you need.
Can you try Anakin without an API key?
Yes. Zero Touch lets you scrape a page or run a read-only Wire action with no key and no signup, metered per IP. When the project is ready, you upgrade in place by adding an X-API-Key header.
Is Google's lawsuit against SerpApi settled?
No. A court dismissed Google's DMCA claims against SerpApi in July 2026, but Google filed a narrower amended complaint in August, and SerpApi has moved to dismiss it again.
The bottom line
It comes back to the page. If everything you need lives on a search results page, SerpApi is a strong tool and its legal posture is a bonus. If you need the page behind the result, rendered, behind a login, or structured to your own schema, that's what Anakin is built for, and you're charged only when the call succeeds.
