Google Academic Scraper — Scholar + Patents

One actor for Google Scholar and Google Patents. No login, no cookies, no API keys — pure HTTP.

What it does

Google Scholar

• Search publications — title, authors, venue, year, snippet, citation count, cited-by URL, "all versions" URL, related-articles URL, PDF link + type, cluster & result ids.
• Author profiles (by Scholar user id) — name, affiliation, verified email, interests, h-index, i10-index, total citations (all / recent), citations-per-year histogram, co-authors, and the full publication list (title, venue, year, citations). Optional derived analytics: citation velocity, recent-2-year citations, h-index cross-check, top venues.
• Cite formats — MLA, APA, Chicago, Harvard, Vancouver + BibTeX / EndNote / RefMan / RefWorks export links.

Google Patents

• Search — free text + filters (assignee, inventor, country, date range by priority/filing/publication, status, type, language, sort). Returns patent number, title, snippet, inventor, assignee, all key dates, PDF, figures, total results available.
• Details — abstract, full claims + count, CPC classifications (with the full scheme/leaf tree), discrete status (GRANT/APPLICATION) and type (PATENT/DESIGN), familyId, forward & backward citations (with totals, assignee, dates, examiner-cited flag), non-patent literature, patent family / worldwide applications, legal status & legal-events timeline, PDF.
• Family and N-hop citation network — each node carries citationHop (graph distance from the seed) and citationSeed (origin patent), with global de-duplication across hops.

Google Patents renders CPC only (no IPC codes exist in its markup for any patent), so this actor ships CPC and does not claim IPC.

Set source to scholar, patents, or both.

Quick start

Scholar search:

{ "source": "scholar", "scholarMode": "search", "queries": ["deep learning"], "maxResults": 20 }

Author profile:

{ "source": "scholar", "scholarMode": "authorProfile", "authorIds": ["WLN3QrAAAAAJ"] }

Patents search:

{ "source": "patents", "patentsMode": "search", "searchQuery": "neural network", "maxResults": 50 }

Patent details (deep):

{ "source": "patents", "patentsMode": "details", "patentIds": ["US10000000B2"], "includeClaims": true }

Output

One dataset item per result, tagged with source (scholar/patents) and type (scholar_search_result, scholar_author_profile, scholar_cite_formats, patents_search_result, patents_details). Blocked/failed requests are emitted as type:"error" (with a blockReason when identifiable) — never silently dropped.

Proxy

Proxies are configured automatically for reliable results — just make sure your Apify plan includes proxy access. Google Scholar throttles heavily, so the actor retries automatically; under very heavy throttling a run may occasionally return fewer results (failed requests are emitted as type:"error", never silently dropped — simply re-run to fill any gaps).

Pricing

Pay-per-event:

Event	Price
Actor start	$0.01 per run
Result	$0.007 per record delivered (Scholar or Patents)

You're charged only for successful records — failed / blocked requests (type:"error") are never billed. Apify platform usage is billed to your own Apify account on top of these events.

Support & feedback

Found a bug, or want another Scholar/Patents field or mode? Open an issue on the actor's Issues tab, or leave a rating and review on the actor page — it genuinely helps and I read every one.

Legal / ToS

This actor scrapes only publicly available Google Scholar and Google Patents pages. It does not bypass any login or paywall and stores no personal data beyond what Google publicly renders. You are responsible for compliance with Google's Terms of Service and applicable law in your jurisdiction.