Google Patents Search Scraper

Google Patents Search Scraper exports public patent search results and detail metadata from queries, Google Patents URLs, or publication IDs. Collect titles, inventors, assignees, dates and source links for prior-art discovery and IP research, then download CSV, Excel or JSON.

Search snapshots and detail lookups have different field coverage. Abstracts, classifications and citations depend on available detail pages; this Actor does not extract full claims or expand patent-family graphs.

Quick start

Open the input, replace the sample queries and clear any unwanted sample patent IDs. Start with an explicit small limit:

{
  "queries": ["assignee:(Tesla) battery"],
  "patentUrls": [],
  "maxItems": 5,
  "includeDetails": true,
  "proxyConfiguration": { "useApifyProxy": false }
}

Run the Actor, then open its dataset to view or export records. Provide at least one query or patent URL/ID. When both are supplied, direct patent inputs are processed first and share the same record limit.

Ready-to-run examples

These existing examples include their own connection settings; check the input before starting.

• Fast Google Patents search export — Save up to 10 search records without detail enrichment.
• Extract metadata from patent IDs — Look up two known publication IDs and their available detail fields.
• Build an EV charging patent dataset — Collect up to 20 results for charging-infrastructure research, with detail enrichment enabled.

Who is it for?

• Patent researchers: create bounded prior-art search samples with source URLs for follow-up review.
• R&D and IP analysts: compare technologies, assignees and inventors using Google Patents query syntax.
• Data teams: join known publication IDs to available abstracts, classifications and citation metadata.

Input configuration

Setting	JSON key	Purpose
Search queries	queries	Search text or operators such as assignee:(Tesla) and inventor:(Smith).
Patent URLs or IDs	patentUrls	Direct publication lookups, processed before queries.
Maximum patent records	maxItems	Total saved-row cap across all inputs; set an explicit value for API calls.
Include detail-page metadata	includeDetails	Add available detail fields to search records; direct lookups always fetch details.
Proxy configuration	proxyConfiguration	Optional connection settings; direct access is the default. See field help for supported settings.

For a known patent, use this alternative input. No search query is needed:

{
  "queries": [],
  "patentUrls": ["https://patents.google.com/patent/US10000000B2/en"],
  "maxItems": 1,
  "includeDetails": true
}

Output fields

Each saved row represents one patent publication. Duplicate publication IDs are saved only once across a run; rank is the position in the saved set, not a separate ranking for each query.

Illustrative shape only, not an observed patent record; the excerpt omits other fields:

{
  "query": "solar panel",
  "rank": 1,
  "patentId": "US10000000B2",
  "patentUrl": "https://patents.google.com/patent/US10000000B2/en",
  "patentTitle": null,
  "abstract": null,
  "classifications": [],
  "citations": [],
  "source": "search",
  "warnings": []
}

Fields	Meaning and availability
patentId, patentUrl, patentTitle	Publication ID, source page and available title.
query, rank	Originating query and saved-set position; null for direct inputs.
snippet, abstract	Search snippet; abstract requires available detail metadata.
inventor, assignee	Search-provided strings; null on direct-only records.
inventors, assignees	Parsed lists; direct lookups use available contributor metadata. Verify important attribution against the source.
priorityDate, filingDate, publicationDate, grantDate	Available date values. Detail-derived grantDate is the last page date, not a verified legal grant date.
applicationNumber, publicationNumber	Available application number and publication identifier.
language, status, jurisdictions	Search-provided language and family-level status/jurisdiction hints; not a legal-status determination.
classifications, citations	Detail-page codes and up to 100 relation values. Not a complete forward/backward citation network.
pdfUrl, thumbnailUrl	Available links, not downloaded PDF/image files.
source, scrapedAt, warnings	search or detail, collection time and any optional-enrichment warnings.

Direct-only records do not gain search-only fields such as snippet, priority/publication dates, language or status. Missing values remain null or empty arrays.

Pricing and limits

The Actor uses pay-per-event pricing: start is a one-time run-start event, and item is charged for each saved patent record, including a search record retained with an enrichment warning. Failed direct lookups do not save empty records. See the live Pricing tab for current rates.

• Bound collection: maxItems limits saved rows from 1 to 10,000; it does not guarantee that many matches.
• Choose detail coverage: explicitly set includeDetails to false for lighter search snapshots, or true for available abstracts, classifications and citations.
• Expect incomplete pages: Google can throttle requests or omit metadata. A failed optional detail lookup can leave the useful search record intact with warnings.
• Review run state: RUN_SUMMARY normally records saved counts, warnings and failures. If time-bounded work remains, PENDING_WORK lists remaining scopes; it is not an automatic resume feature. These records may be unavailable if storage fails.
• Interpret empty results: a valid search with no matches can return an empty dataset. If every attempted source fails and no record is saved, the run fails. Consult the summary and report unexpected failures.
• Verify legal conclusions: this is public metadata for research, not an official patent-office service or legal advice.

API usage

Use the same bounded inputs from Node.js, Python or HTTP. Read the dataset after a successful run.

Node.js

import { ApifyClient } from 'apify-client';
const client = new ApifyClient({ token: process.env.APIFY_TOKEN });
const run = await client.actor('fetch_cat/google-patents-search-scraper').call({
  queries: ['solar panel'], patentUrls: [], maxItems: 5, includeDetails: true
});
console.log((await client.dataset(run.defaultDatasetId).listItems()).items);

Python

import os
from apify_client import ApifyClient
client = ApifyClient(os.environ['APIFY_TOKEN'])
run = client.actor('fetch_cat/google-patents-search-scraper').call(run_input={
    'queries': ['solar panel'], 'patentUrls': [], 'maxItems': 5, 'includeDetails': True
})
print(client.dataset(run['defaultDatasetId']).list_items().items)

HTTP

curl -X POST 'https://api.apify.com/v2/acts/fetch_cat~google-patents-search-scraper/runs' \
  -H "Authorization: Bearer $APIFY_TOKEN" \
  -H 'Content-Type: application/json' \
  -d '{"queries":["solar panel"],"patentUrls":[],"maxItems":5,"includeDetails":true}'

The HTTP endpoint starts an asynchronous run. Wait for completion and use the returned defaultDatasetId to retrieve results.

MCP and AI agents

Use the official Apify MCP server, not an Actor-specific server. Its focused URL is:

https://mcp.apify.com?tools=fetch_cat/google-patents-search-scraper

Add it as an HTTP MCP server in a compatible client and authenticate with Apify. The general server can discover Actors; the focused URL selects this Actor. Ask: “Run fetch_cat/google-patents-search-scraper with queries: [\"solar panel\"], patentUrls: [], maxItems: 5, includeDetails: true; return the dataset IDs, titles and source links.” Treat null fields as unavailable, not evidence of no inventor or no patent rights.

Related Actors

• Google Scholar Profiles Scraper — Research author profiles alongside patent research.
• Google News Scraper — Add company and technology news context.
• Google Autocomplete Scraper — Explore related search phrases.
• Google Search Results Scraper — Find supporting public web sources.
• Google Trends Scraper — Compare public search-interest context for technologies.

FAQ

Can I search Google Patents by inventor or assignee? Yes. Put Google Patents syntax such as inventor:(Smith) or assignee:(Tesla) in queries. Check the returned source pages for relevance.

Is this a Google Patents search API? It is an independent Apify workflow for collecting public Google Patents metadata, callable through API or MCP. It is not an official Google API and requires no Google login or API key.

Can I export a patent dataset to CSV or Excel? Yes. Open the run's dataset and select an export format, or retrieve JSON through the dataset API.

Why are abstracts or citations missing? Search-only records do not contain these detail fields. Set includeDetails: true; available detail pages can still omit values. Check row warnings and source pages before drawing conclusions.

Can I extract full claims or patent-family graphs? No. The output is search/detail metadata, with limited citation values rather than graph traversal. For broader tools, compare their supported fields before choosing an alternative.

Support and feedback

For custom additions or integrations, email hello@fetchcat.dev.

Found an error or have a feature request? Open an issue with the run ID, a non-sensitive input example and what you expected.

Made with care for curious people, AI agents and the possibilities in data. ♥