Google Scholar API: what exists, what it costs, what it risks
There is no official Google Scholar API, and Scholar's robots.txt disallows /scholar for every crawler. A Google Scholar API is therefore a third party reading the results page: 6 providers here return Scholar results as JSON, from $1.00 per 1,000 requests on Serper at 10,000 a month. If you only need paper metadata and citations, open scholarly databases are a cleaner source.
Reviewed by Eugene SuslovFacts read November 2026No affiliate links

- 6
- providers with a Scholar endpoint, priced
- $1.00
- per 1,000 at 10,000 a month, the cheapest
- None
- official API from Google
- 1,000
- results Scholar shows for any one query, in Google's words
What a Scholar result holds
Each result is a paper as Scholar indexes it: title, authors, publication and year, a snippet, a link to the publisher page and, where Scholar found one, a PDF. Underneath sit the fields researchers actually want: the cited-by count with a link to the citing papers, the number of versions, and related articles.
Scholar's own outputs stop there. It exports one citation at a time to BibTeX, EndNote, RefMan or RefWorks, and sends email alerts. Anything at scale, from a literature review to tracking who cites a lab's papers, needs a Scholar endpoint from a third party.
SERP APIs with a Google Scholar endpoint
6 providers document a structured Scholar endpoint. Each card links to the full review with plans, limits and terms.
Showing 6 APIs

Best documented SERP API
The most thoroughly documented SERP API, covering Google AI Mode and AI Overviews alongside dozens of engines. It indemnifies US scraping claims up to $2m from the Production plan, and is itself in court with Google.
- Dozens of engines including Google AI Mode and AI Overview
- Legal indemnification up to $2m on Production and above
- Throughput measured per hour, not per second
- SERP results
- Local & maps
- Content & SERP features

Cheapest credible SERP data
Prepaid Google search credits with sub-two-second responses and no subscription. The cheapest credible SERP data here at volume, though the entry rate is $1 per 1,000, not the headline $0.30.
- Search, Images, News, Maps, Places, Videos, Shopping, Scholar
- Prepaid credit packs, no monthly commitment
- Credits deducted only on successful responses
- SERP results
- Local & maps
- Content & SERP features

Best structured SERP parsing
Structured SERP parsing on a monthly plan whose fee becomes credit, spent at a per-target rate. Google is $1.00 per 1,000 results on the $15 Starter plan, falling to $0.50 at Enterprise.
- Per-target pricing: Google rated separately from other sites
- Monthly plans from $15, spent as credit at per-target rates
- JavaScript rendering priced as a separate, higher rate
- SERP results
- Local & maps
- Content & SERP features

Best budget multi-engine SERP API
A budget SERP API with dedicated endpoints for Google and its AI Overviews and AI Mode, plus Bing, Baidu, Yahoo and DuckDuckGo. Plans start at $40 a month, and only Standard and above carry unused credits over.
- Google Search API plus AI Overview and AI Mode endpoints
- Bing, Baidu, Yahoo, DuckDuckGo and YouTube endpoints
- LinkedIn, Amazon and general web scraping
- SERP results
- Content & SERP features
- AI search & LLM visibility

Best AI engine coverage
A SERP API that covers ChatGPT Search and Perplexity alongside Google, Bing and a long list of other engines, which makes it one of the more practical ways to monitor AI answer engines through a normal search endpoint.
- Google, Bing, YouTube, ChatGPT Search and Perplexity
- Structured JSON across every engine
- AI engine results through a standard SERP interface
- SERP results
- AI search & LLM visibility
- Content & SERP features

Best for very large batches
ValueSERP's sibling from Traject Data, with more Google search types (Scholar, Autocomplete, Places detail) and a batch endpoint. It costs more per search than ValueSERP at every volume we can compare.
- Google search, news, places, shopping, images and Scholar
- Batch endpoint built for large keyword sets
- AI Overviews via a parameter on web search
- SERP results
- Local & maps
- Content & SERP features
What Scholar requests cost on each provider
Enter your monthly requests. Each provider is priced on its cheapest published route.
- 1Serper50k pack + overage$10.00$1.00 / 1k
- 2Oxylabs SERP Scraper APIStarter · Needs JavaScript rendering, billed at the JS rate$15.00$1.50 / 1k
- 3ScrapingdogLite · 5 credits a request$40.00$4.00 / 1k
- 4SearchApiDeveloper · Up to 20 results per request$40.00$4.00 / 1k
- 5Scale SERP10,000$83.00$8.30 / 1k
- 6SerpApiProduction · Up to 20 results per request$150$15.00 / 1k
Cheapest published route for that volume, monthly billing, before tax. A vertical request is priced on the same plans as the provider's Google searches, scaled by its own credit cost. Prices read in November 2026.
At 10,000 requests a month Serper is cheapest at $10.00 and SerpApi costs $150; at 100,000 the order is Serper, Scrapingdog and SearchApi. Scrapingdog bills a Scholar request at half a Google search. Oxylabs SERP Scraper API bills more, because the page needs JavaScript rendering.
SerpApi and SearchApi return up to 20 results per request against the usual 10, which halves the requests for a deep query. A cited-by list is paged like a search, so tracking citations to one paper costs a request per page of citing papers.
No API, and a robots.txt that says no
Scholar is the strictest Google property on automated access, in its own words.
Scholar's robots.txt tells every crawler "Disallow: /scholar", and its help pages refuse both bulk access and higher limits:
“Sorry, we're unable to provide bulk access. You'll need to make an arrangement directly with the source of the data you're interested in.”
Google Scholar help, read 20 November 2026
Google's Terms of Service list among the abuses "using automated means to access content from any of our services in violation of the machine-readable instructions on our web pages (for example, robots.txt files that disallow crawling, training, or other activities)". That is a contract term, not a court ruling, and the providers here take on the collection; but it means a product built on Scholar data carries more terms risk than one built on web search results. Read each provider's terms and ask whether it indemnifies you.
The 1,000-result cap, and how to work inside it
Scholar never shows more than 1,000 results for one query, through any endpoint.
Google's help says it can only show up to 1,000 results for a query, so no endpoint can page past it. A broad query is not a way to download a field.
- Split by year. The same query, one year at a time, gives each year its own 1,000.
- Add a source or author filter to narrow a crowded topic.
- Go through cited-by from a few key papers rather than through a keyword, when a literature map is the goal.
Author profiles: the Scholar pages robots.txt allows
The same robots.txt that disallows search makes an exception that matters for citation tracking.
Scholar's robots.txt disallows search pages and most citation pages, but allows public author profiles (the /citations?user= pages) and the top-venues lists. A profile carries an author's papers, citation counts by year, h-index and i10-index.
So tracking a researcher, a lab or a department through their profiles reads pages Google lets crawlers fetch, which is a cleaner position than querying search results at volume. It only works for authors who made a public profile, and the counts are Scholar's own, which run higher than curated databases because Scholar indexes more sources.
Check whether a provider has a dedicated author endpoint before you choose: an author request is a different call from a search, and not every Scholar endpoint offers one.
What teams use Scholar data for
Four jobs, and which part of Scholar each one reads.
- Literature reviews. Queries by topic and year, ranked as Scholar ranks them; the 1,000-result cap is the constraint to plan around.
- Citation monitoring. The cited-by lists of a set of papers, checked monthly, to see who builds on a lab's work.
- Academic visibility for publishers. Where a journal's papers rank on Scholar for the topics it covers, the scholarly version of rank tracking.
- Research for products. Edtech, legal and life-science tools that point users at papers, usually with an open database for metadata and Scholar for ranking.
The first two read search and cited-by pages, the ones robots.txt disallows. The third reads search pages for a fixed set of queries, which keeps volume low. Size each one by requests a month before you pick a plan.
One key for Scholar and the rest of Google
Scholar is a side endpoint for every provider here.
| Provider | Also returns on the same key |
|---|---|
| Serper | Google webImagesLensNewsMapsShopping |
| Oxylabs SERP Scraper API | Google webAI OverviewsTrendsImagesLensNewsShoppingYouTubeBing |
| Scrapingdog | Google webAI OverviewsTrendsImagesLensNewsMapsShoppingYouTubeBing |
| SearchApi | Google webAI OverviewsTrendsImagesLensReverse imageNewsMapsShoppingYouTubeBing |
| Scale SERP | Google webAI OverviewsImagesNewsMapsShopping |
| SerpApi | Google webAI OverviewsTrendsImagesLensReverse imageNewsMapsShoppingYouTubeBing |
Engines each provider returns as structured JSON, read from its docs on 20 November 2026.
Serper, Oxylabs SERP Scraper API, Scrapingdog, SearchApi, Scale SERP and SerpApi also return Google News, the pairing for tracking how research is covered in the press. None of them sells Scholar alone, so the price you pay is set by the plan you would buy for Google search.
Free allowances for Scholar requests
Enough to test a provider on your hardest query.
| Provider | Free Scholar requests | How often |
|---|---|---|
| SerpApi | 250 | Every month |
| Scale SERP | 125 | Every month |
| Oxylabs SERP Scraper API | 3,703 | Once, at sign-up |
| Serper | 2,500 | Once, at sign-up |
| SearchApi | 100 | Once, at sign-up |
| Scrapingdog | 20 | Once, at sign-up |
Each provider's free allowance converted at its own Scholar credit cost.
Test with a query that has a long cited-by list and a few hundred results: it shows how each provider handles paging, versions and missing PDFs, which is where Scholar endpoints differ.
When an open scholarly database is the better source
Scholar's value is its ranking and its reach into grey literature. Metadata alone is available elsewhere.
Open scholarly databases such as OpenAlex, Crossref and Semantic Scholar publish their own APIs for paper metadata and citations, with bulk access the publishers allow. For counting papers, mapping citations or enriching a reference list, start there: there is no robots.txt to work against and nothing to scrape.
Use a Scholar endpoint when what you need is what Scholar shows: its ranking for a query, its version clusters, or papers and theses the open databases miss. Many projects use both, with Scholar as the check on coverage.
What a citation-tracking job costs
A worked example: a lab tracking who cites its papers, every month.
Say 200 papers, each with a cited-by list averaging 3 pages of results, checked monthly: 600 requests a month. Add a monthly search for 50 topic queries at two pages each, and the job is 700 requests.
| Provider | 700 requests a month | Route |
|---|---|---|
| Serper | $8.33 | 50k pack |
| Oxylabs SERP Scraper API | $15.00 | Starter |
| SerpApi | $25.00 | Starter |
| Scale SERP | $29.00 | 1,000 |
| Scrapingdog | $40.00 | Lite |
| SearchApi | $40.00 | Developer |
Priced on each provider's cheapest published route, November 2026.
At this size the plan floor decides more than the rate: Serper covers it for $8.33, while the dearest route costs $40.00. Twenty results per request, where offered, would roughly halve the cited-by part.
The Scholar endpoints side by side
Price per 1,000 requests at 10,000 a month.
| Provider | Per 1,000 | Route | Worth knowing |
|---|---|---|---|
| Serper | $1.00 | 50k pack + overage | Same price as a Google search |
| Oxylabs SERP Scraper API | $1.50 | Starter | Needs JavaScript rendering, billed at the JS rate |
| Scrapingdog | $4.00 | Lite | 5 credits a request |
| SearchApi | $4.00 | Developer | Up to 20 results per request |
| Scale SERP | $8.30 | 10,000 | Same price as a Google search |
| SerpApi | $15.00 | Production | Up to 20 results per request |
Read from each vendor's pricing page and docs in November 2026.
How to choose a Scholar provider
Four checks, in order.
- 01Do you need Scholar, or the metadata?If the metadata, use an open database's API first.
- 02Results per requestTwenty per request halves the cost of a deep query.
- 03Cited-by and author profilesConfirm the endpoint parses the cited-by link and the fields you track.
- 04Terms and indemnityScholar is the riskiest Google property to build on; read the provider's terms.
Google Scholar APIs, answered
Is there an official Google Scholar API?
No. Google offers no Scholar API, declines bulk access, and asks automated software to respect a robots.txt that disallows /scholar. Every endpoint here is a third party.Is it allowed to use a Google Scholar API?
Google's terms prohibit automated access that ignores robots.txt, and Scholar's robots.txt disallows search pages. That is a contract matter between Google and whoever collects the data; read your provider's terms before you build a product on it.How many results can a Google Scholar API return for one query?
At most 1,000, because Scholar shows no more for any query. Split by year or filter by source to go further.Can I get Scholar author profiles and h-index by API?
Yes, where a provider has an author endpoint: public profiles carry papers, citations by year, h-index and i10-index. They are also the Scholar pages robots.txt allows, which makes them the lowest-risk part of Scholar to read.Do Scholar endpoints return PDF links and citation counts?
They return what the results page shows: a PDF or HTML link where Scholar found one, the cited-by count with a link to the citing papers, and the number of versions. A paper with no free copy has no PDF link on any endpoint.What is the cheapest way to get Scholar results by API?
At 10,000 requests a month, Serper at $10.00, then Oxylabs SERP Scraper API and Scrapingdog.
Picking the provider is the easy half.
busyless runs content and search as a system: we choose the data sources and tools, build the pipelines on top, and measure what they feed. If that is the part you are missing, talk to us.
Ask an assistant about this page
An independent comparison. No provider pays to be listed or to rank higher, outbound links carry nofollow, and we earn nothing if you sign up. Prices and terms were read from each vendor's own pages in November 2026 and change without notice.
Missing, or out of date?
Every entry here was read off the company's own website and dated, so some of it will have moved since. Corrections are free and so is being added. There is no fee, no sponsored position, and nothing that money moves up this page.