Turn a Google search into clean JSON. Thirteen surfaces, one request shape. Credits never expire, and an empty result is free.
{
"searchParameters": {
"q": "best running shoes",
"gl": "us",
"hl": "en",
"type": "web",
"engine": "google"
},
"organic": [
{
"title": "The 15 Best Running Shoes of 2026",
"link": "https://www.runnersworld.com/gear/a19663621/best-running-shoes/",
"snippet": "For new runners we recommend a shoe like the Brooks Ghost or Nike Pegasus a...",
"position": 1
},
{
"title": "What are the best running shoes that you'd actually buy ...",
"link": "https://www.reddit.com/r/runninglifestyle/comments/1u5x7mx/what_are_the_best_running_shoes_that_youd/",
"snippet": "I\u2019m finally replacing my running shoes and I forgot how overwhelming this c...",
"position": 2
},
{
"title": "100+ Running Shoe Reviews",
"link": "https://runrepeat.com/catalog/running-shoes",
"snippet": "Running shoe reviews \u00b7 Nike Vomero Plus. 91. Superb. Nike Vomero Plus \u00b7 adi...",
"position": 3
}
],
"peopleAlsoAsk": [
{
"question": "What are the top 5 best running shoes?"
},
{
"question": "What are the best shoes for running?"
}
],
"tookMs": 931
}Thirteen Google surfaces behind one request shape. Tap a surface below and watch the request change.
Credits expire in a few months. Empty pages bill the same as full ones. Go read the reviews. The first page is all expiry complaints.
A search that returns results costs one credit. A search that fails costs nothing. What you buy sits there until you use it.
SERP stands for search engine results page. A SERP API fetches that page from Google at scale and returns it parsed: organic results with positions, the local pack, knowledge graph, news, shopping, images, reviews.
You send a query over HTTPS with an API key and get structured data back, usually in a second or two. The scraping and parser upkeep are the vendor's problem instead of yours.
Who runs on these: rank trackers checking positions daily, lead-gen tools pulling business lists from Google Maps, price monitors watching shopping results, research pipelines, and AI agents that need live search grounding. If your product needs Google results as data, this is the category you're shopping.
On the legal side: logged-out, public-only collection has repeatedly defeated computer-fraud and contract claims, while Google's Terms prohibit automated access. Vendors carry that tension so buyers don't have to.
Every vendor returns Google results. The differences hide in the billing fine print and the schema. Check these five:
Per request, per result, or per successful result? The sharpest question: what happens when Google returns nothing? On most vendors an empty page still bills. At long-tail volume that quietly becomes a real line item.
Some vendors reset unused searches every month. Others expire prepaid credits after a set window. Read that clause before buying a big pack, because "unused" is the normal state of a prepaid balance.
You will write parsers against this schema and live with them. Check that the shape is stable, documented, and consistent across search types. The serper-style JSON shape has become the de facto standard, and we kept ours compatible with it on purpose.
Rank tracking needs web. Lead gen needs maps and reviews. Monitoring needs news and shopping. Count the search types on your roadmap, then check the vendor actually serves them as parsed data.
Anything over a couple of seconds hurts agents and user-facing features. Every serious vendor has a free tier. Use it: run your real queries and time them before any money moves.
Three names come up most when devs shop this category: SerpApi, Serper, and us. All three are good at their core job, and pretending otherwise would be silly. Here's the plain version:
| SerpApi | Serper | CrustAPI | |
|---|---|---|---|
| Known for | The veteran. Deep parsing, biggest catalog | Cheap, fast Google credits | Bills only on results, credits never expire |
| Engines beyond Google | Yes, many | No, Google only | No Bing or Baidu. Google plus LinkedIn public data |
| Billing model | Monthly subscription | Prepaid credits | Prepaid credits, 1 per query that returns results |
| Empty results | Billed | Billed | Free |
| Do unused credits expire? | Searches reset monthly | Yes | Never |
| Free tier | Yes | Yes | 3,000 / month, no card |
| Legal cover product | Sold on bigger plans | No | No |
The prices move, so we keep receipt-backed vendor tables current instead of hardcoding them here: SerpApi vs Serper vs CrustAPI is the direct three-way, SERP API pricing maps the whole category's cost per 1,000 in one table, and the Serper roundup and SerpApi roundup rank ten vendors each on real per-1,000 pricing and billing quirks. For the flagship surface on its own, see the Google Search API page.
The trade-offs, since you're comparing vendors: CrustAPI is a one-person company. Dawson builds it and answers support@crustapi.com himself, so you get direct answers and no sales calls. There are no compliance certifications or enterprise account team; if procurement needs those, SerpApi is the safer pick.
On CrustAPI, 1 credit buys 1 search that returns results; Maps instead bills 1 credit per business returned. Empty and failed calls cost nothing.
| Pack | You pay | Per 1,000 searches |
|---|---|---|
| 3,000 credits every month | Free, no card | $0 |
| 25,000 credits | $49 | $1.96 |
| 100,000 credits | $149 | $1.49 |
| 500,000 credits | $549 | $1.10 |
| 2.5 million credits | $1,999 | $0.80 |
Packs continue up to 250 million credits for $100,000, which is $0.40 per 1,000. Prices ex-tax.
Credits never expire. Try your real workload in the playground or straight from the docs.
Create a free account. You get 3,000 credits a month, no card needed.
Run a search in the playground. Pick any of the 13 types and type a query. Results land in a table.
Copy your key and run any curl on this page, or export the results as CSV.
The same account also covers LinkedIn public data: full profiles, companies, jobs, and posts, with verified emails on people results. It pairs naturally with the SERP side. Maps records carry a business's name, address, phone, website, ratings, and hours; when you need a decision maker's verified email, that comes from the LinkedIn people endpoints.
When someone else is the better pick: SerpApi if you need engines beyond Google (Bing, Baidu, and many more) or your legal team wants the liability cover they sell on bigger plans. Serper if you want the cheapest entry price for Google credits and you'll reliably burn your volume before it expires. Both are solid products run by people who know this space well.
SERP stands for search engine results page. A SERP API fetches live Google results and returns them as structured JSON: organic results with positions, the local pack, knowledge graph, news, shopping, images. It replaces running your own proxies, headless browsers, and parsers.
Both, which trips people up. SerpApi (serpapi.com) is a vendor that named itself after the category. The generic term is SERP API, and there are many vendors, including Serper and CrustAPI. If you searched 'serp api' looking for the company, you want serpapi.com. If you're shopping the category, keep reading the comparisons.
Stated neutrally: logged-out, public-only collection has repeatedly defeated computer-fraud and contract claims, while Google's Terms prohibit automated access. SERP API vendors operate in that space so buyers don't have to, and some sell legal cover on bigger plans.
It varies a lot by vendor and volume. CrustAPI charges 1 credit per query that returns results: $49 buys 25,000 and $1,999 buys 2.5 million, with 3,000 free every month and credits that never expire. For vendor-by-vendor price tables with receipts, see crustapi.com/serpapi-vs-serper and the roundup pages.
Because a real share of at-scale queries return nothing: long-tail keywords, small-town lead sweeps, exploratory agent calls. Per-request billing charges you for all of them. CrustAPI charges only for queries that return results, so dead ends cost nothing.
No. That's the whole point of the category. You call one HTTPS endpoint with an x-api-key header and parse JSON. Fetching, blocking, and parser upkeep are the vendor's job.
