Skip to content

Serper vs SerpApi: Google SERP APIs Compared (2026)

Serper vs SerpApi, compared on how you pay, what you can search, free allowances and what a small month really costs. Plus a third option that starts free every month, with a real run, code and the MCP setup for AI agents.

Published
Reading time
8 min read
Pixel-art robot with an orange >_ on its terminal head walks past two switched-off arcade cabinets bearing the Serper and SerpApi logos and rests its hand on a lit cabinet bearing the Scrape.ly logo, thinking of the Scrape.ly mark
In this article
  1. Serper vs SerpApi at a glance
  2. Pricing: credit packs vs subscriptions
  3. Serper
  4. SerpApi
  5. Which one is cheaper?
  6. Coverage and features
  7. The third option in practice: a real run
  8. Which should you choose?
  9. What to watch out for with any SERP API
  10. FAQ
  11. Is Serper cheaper than SerpApi?
  12. What is the Serper API?
  13. Is SerpApi free?
  14. What is the best Serper alternative?
  15. Which is better for AI agents, Serper or SerpApi?

Serper vs SerpApi comes down to how you buy and what you search. Serper sells prepaid credit packs for Google search, from $50 for 50,000 credits that last six months. SerpApi sells monthly subscriptions, from $25 for 1,000 searches, and covers Google plus dozens of other engines.

Both are good at returning search results as JSON. Both also ask you to commit money before your first real month: a pack or a plan. For most teams, the better buy is a third option, Scrape.ly's Google Search Scraper, which charges per results page and starts free with $5 of usage every month; more usage comes with the $19 Starter plan, all of it usage credit. This guide compares all three. Every Serper and SerpApi fact below was checked on 27 October 2026 on serper.dev and serpapi.com/pricing.

Serper vs SerpApi at a glance

Serper SerpApi Scrape.ly
How you pay Prepaid credit packs Monthly subscription Per results page, from your balance
Smallest purchase $50 pack (50,000 credits) $25 a month (1,000 searches) Nothing: start on the free credit; Starter is $19 when you need more
Free allowance 2,500 free queries, no card 250 searches a month $5 of credit every month, no card
Allowance clock Credits valid for 6 months Searches reset each billing cycle The free $5 comes back every month
Search engines Google (search, images, news, maps, places, videos, shopping, scholar, patents, autocomplete) Google plus Bing, Baidu, Amazon, Apple App Store and more Google and Bing, plus 560+ scrapers for other sites
Spending cap Stops when credits run out Plan allowance, with optional early renewal A maximum charge you set on every run

Pricing: credit packs vs subscriptions

Serper

Serper's pricing is a top-up model: "No monthly subscriptions; simply pay for the credits you need." The packs on its site:

Pack Price Credits Queries per second
Starter $50 50,000 50
Standard $375 500,000 100
Scale $1,250 2,500,000 200
Ultimate $3,750 12,500,000 300

Prices exclude sales tax and VAT, and "Credits valid for 6 months". Credits are deducted only for successful responses, and "when your credit balance reaches zero, we will stop accepting new queries." New accounts get 2,500 free queries without a card. Serper offers a full refund within 7 days of your first purchase "unless you have used more than 20% of your credits."

SerpApi

SerpApi sells monthly plans with a number of searches and an hourly throughput:

Plan Price a month Searches a month Throughput an hour
Free $0 250 50
Starter $25 1,000 200
Developer $75 5,000 1,000
Production $150 15,000 3,000
Big Data $275 30,000 6,000
Searcher $725 100,000 20,000
Volume $1,475 250,000 50,000

Larger Cloud plans run from 1 million searches ($3,750) to 54 million a month. "Only successful searches are counted", and "cached, errored, and failed searches are not." An empty response counts the same as one with 100 results. If you run out mid-month, you can renew early, which replenishes your searches and moves your renewal date. SerpApi also refunds within 7 days unless you've used more than 20% of your searches. Our SerpApi pricing breakdown works through every plan.

Which one is cheaper?

Per search at volume, Serper's packs cost less than SerpApi's plans, while SerpApi's plans cover engines and features Serper doesn't. The real question for most teams is how much they have to buy up front. At a few hundred or a few thousand searches a month, the smallest purchase dominates the bill:

1,000 Google searches a month Serper SerpApi Scrape.ly
First month Covered by the 2,500 free queries $25 (Starter) $3.30, covered by the free $5 credit
After that A $50 pack, used within 6 months $25 every month $3.30, still covered by the free credit

The arithmetic for Scrape.ly: 1,000 pages × $0.0033 = $3.30, under the $5 the Free plan adds every month. Over a year of 1,000 searches a month, that's $0 on Scrape.ly's free credit, $300 on SerpApi's Starter plan, and at least one $50 pack on Serper.

Coverage and features

Serper is a Google search API. Its site lists endpoints for search, images, news, maps, places, videos, shopping, scholar, patents and autocomplete, with location set by country and language. It says queries "are returned in 1 to 2 seconds" and that "we do not cache anything", and it lists integrations with LangChain, CrewAI, Haystack and Jan AI.

SerpApi covers many engines and Google result types. Its site lists APIs for Google Search, AI Mode, AI Overviews, Flights, Hotels, Jobs, Lens and more, plus Bing, Baidu, Amazon and the Apple App Store. It lists U.S. Legal Shield on its plans, which "provides up to $2 million in coverage for the scraping and parsing of search engine data", a ZeroTrace Mode that stores no search data, and SOC 2 Type II certification.

Scrape.ly gives you Google results through the Google Search Scraper, one JSON item per results page with organicResults (position, title, URL, displayed URL, snippet, date), peopleAlsoAsk, relatedQueries and hasNextPage. Query operators like site:, intitle:, file types and date ranges are input fields. Turn on live AI Overviews or ads and the run switches to browser mode for those pages. And because it's one Actor in a Store of 560+, the same account crawls the ranking pages with the Website Content Crawler, searches Bing with the Bing Search Scraper and tracks AI answers with the Google AI Overviews Scraper.

The third option in practice: a real run

We ran the Google Search Scraper on 27 October 2026 with five US queries, one page each, in fast search mode: "serper vs serpapi" and four others we were researching. The run returned 5 pages with 50 organic results in 5 seconds, at a list price of 5 × $0.0033 = $0.0165 (our account was charged $0.0132 because its plan has a Store discount). Here's who held the top five spots for "serper vs serpapi":

Position Domain Title
1 serpapi.com SerpApi VS Serper VS ValueSERP VS SearchApi (2026)
2 medium.com SerpAPI vs Serper vs Scrapingdog: We Tested All Three …
3 serper.dev Serper - The World's Fastest and Cheapest Google Search API
4 scrapingdog.com 10 Best SERP APIs in 2026 to Scale Data Extraction from …
5 reddit.com We should look for another option besides SerpAPI, right?

The same page also returned Google's related searches, including "Serper vs Tavily", "Serper API key" and "Serper alternative". The log line that matters:

Text
2026-10-27T16:02:13.329Z [scrapely.actor] INFO  Pushed 5 dataset items, 50 organic results, 0 failed pages in 0.8s. Pages handled in fast search mode: 5, in browser mode: 0.

A run this short fits the synchronous endpoint, which returns the items directly:

shell
curl -X POST "https://api.scrape.ly/v2/actors/querystone~google-search-scraper/run-sync-get-dataset-items" \
  -H "Authorization: Bearer $SCRAPELY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"queries": "serper vs serpapi", "maxPagesPerQuery": 1, "countryCode": "us", "languageCode": "en"}'

In Python (pip install scrapely-client-python):

python
import os
from scrapely_client import ScrapelyClient

client = ScrapelyClient(os.environ["SCRAPELY_TOKEN"], api_url="https://api.scrape.ly")

run = client.actor("querystone/google-search-scraper").call(
    run_input={"queries": "serper vs serpapi", "maxPagesPerQuery": 1, "countryCode": "us", "languageCode": "en"}
)
for page in client.dataset(run.default_dataset_id).iterate_items():
    for result in page.get("organicResults", []):
        print(page["searchQuery"]["term"], result["position"], result["url"])

If you're coming from either API, the fields map directly: position, title, link or url, and snippet become organicResults[].position, .title, .url and .description. The Google Search API page has the full field list.

For AI agents, add the hosted MCP server limited to this Actor, https://mcp.scrape.ly?actors=querystone/google-search-scraper, for example in Claude Code with claude mcp add --transport http scrapely https://mcp.scrape.ly. Then ask: "Search Google in the US for 'serp api' and give me the top 10 with position, title and URL." Our list of web search MCP servers covers every client.

Which should you choose?

  • Choose Scrape.ly if your volume is small, uneven or still growing, if you want People Also Ask, related searches and AI Overviews in the same JSON, or if search is one step in a bigger job: crawl the pages that rank, track AI answers, pull Google Maps data or let an AI agent run it all. You start on the free $5 monthly credit, pay $3.30 per 1,000 pages with platform usage included, and set a maximum charge on every run. It's our pick for most teams.
  • Choose Serper if you run a high, steady Google-only volume and will use a $50-or-larger pack within six months.
  • Choose SerpApi if you need many engines beyond Google on one subscription and your monthly volume reliably fills a plan.

For the wider field, our guide to the best SERP APIs compares seven options, and the Scrape.ly vs SerpApi comparison goes feature by feature.

What to watch out for with any SERP API

  • Results differ by country and language. Always send both. Scrape.ly records the query, country and language on every item, so a daily schedule gives you a clean ranking history.
  • Results per page. Google pages hold 10 results. Asking for more means more pages, and every API bills for those somehow: check how before you set a high number.
  • Allowances you don't use. Packs expire and plans reset, so size the purchase to your real volume. On Scrape.ly, small volumes fit inside the free credit that comes back every month.
  • Runaway loops. A bug that repeats a query thousands of times costs real money. Scrape.ly checks the run's budget before each page and stops cleanly at the maximum charge you set.

FAQ

Is Serper cheaper than SerpApi?

Per search, yes: Serper's smallest pack is $50 for 50,000 credits, while SerpApi's Starter plan is $25 for 1,000 searches a month. But Serper requires a $50 purchase after its free queries, and SerpApi a monthly fee. Up to about 1,500 searches a month, Scrape.ly's free $5 monthly credit means you pay nothing at all.

What is the Serper API?

Serper is a Google Search API from serper.dev. You send a query and get Google results as JSON, paying from prepaid credit packs that are valid for six months. It covers Google search, images, news, maps, places, videos, shopping, scholar, patents and autocomplete.

Is SerpApi free?

SerpApi has a free plan with 250 searches a month and 50 searches an hour. Paid plans start at $25 a month for 1,000 searches.

What is the best Serper alternative?

Scrape.ly's Google Search Scraper, for most teams: per-page pricing, a free $5 credit every month, People Also Ask and related searches in every result, and 560+ other scrapers plus an MCP server on the same account. If you're tracking rankings, our guide to SEO rank tracking with scraping shows the full setup.

Which is better for AI agents, Serper or SerpApi?

Both return JSON an agent can read through their APIs. Scrape.ly adds a hosted MCP server that exposes the Google Search Scraper and 560+ other scrapers, so Claude, Cursor or ChatGPT can search Google and then read the pages behind the results through one connection. See how to scrape Google search results for every field.

  • serper
  • serpapi
  • serp api
  • google search
  • comparison

About the author

The team behind Scrape.ly

Scrape.ly is the web data platform from TeraShift GmbH in Switzerland, powered by Evomi: ready-made scrapers in the Scrape.ly Store, a cloud runtime, an API with JavaScript and Python SDKs, proxies, a hosted anti-bot browser and an MCP server for AI agents.

//Keep reading\\

The internet is full of data.
Start owning yours.

Scrape.ly gives you the infrastructure, the tooling, and the team to extract, process, and operationalize web data at any scale. Setup takes minutes.

No credit card. No setup calls.