Google Patents Scraper

📜 Google Patents Scraper

Export patents from Google Patents by keyword, company, inventor or search link, or pull the full record of any patent number: abstract, claims, CPC codes, citations, cited-by, family members, legal status and expiry date. It covers the US, European, WIPO, Chinese, Japanese, Korean and 100+ other patent offices that Google Patents indexes. No login, no API key, and 10 search results usually arrive in about 6 seconds.

✨ Why use Google Patents Scraper?

• ⚡ Fast — search pages come straight from Google Patents' own data feed: 10,000 search results took about 90 seconds in our tests, 50 full records 14 seconds.
• 🔎 Every way in — keywords (with OR, quotes and wildcards), assignees, inventors, pasted search links or plain patent numbers, all in one run.
• 📚 Full records on demand — claims and independent claims, CPC codes with descriptions, backward and forward citations, non-patent literature, family members, legal events, expiry date and litigation count.
• 🕸️ Family and citation crawls — expand a patent to its whole family, or follow citations backward (prior art) and forward (cited by) up to 3 hops.
• 🧮 Beyond 1,000 results — Google Patents shows 1,000 results per search; the scraper splits big searches into date ranges and keeps going, without duplicates.
• 💸 Low price — $1.50 per 1,000 search results and $4.00 per 1,000 full records.

🎯 Use cases

• 🧑‍⚖️ Prior-art search — pull every patent around an invention idea with its claims and citations, then rank by citations or dates.
• 🏢 Competitor and portfolio tracking — list everything an assignee filed, where it is protected and which patents are still active.
• 🗺️ IP landscaping — export thousands of patents in a technology field by office, date range and status for charts and clustering.
• 🧪 R&D and freedom-to-operate research — read the independent claims and expiry dates of the patents closest to your product.
• 👩‍🔬 Inventor and talent research — find the patents of a researcher or inventor across companies.
• 🤖 AI and data pipelines — feed abstracts, claims and CPC codes into search indexes, RAG systems or classifiers.

🚀 How to use

1. Enter one or more search queries, assignees, inventors, Google Patents search links or patent numbers.
2. Set Max results per search (20 by default) and any filters: patent offices, date range, status, patent type, language, litigation, sort order.
3. Turn on Full record for search results if you need claims, citations and family for each search hit. Patent numbers always get the full record.
4. Click Start and watch the results arrive.
5. Download the dataset as JSON, CSV, Excel, XML or HTML, or read it through the API.

📥 Input

{
  "searchQueries": ["solid state battery"],
  "countries": ["US", "EP"],
  "status": "grant",
  "dateFrom": "2020-01-01",
  "maxResults": 100,
  "includeDetails": true
}

A citation crawl from one patent:

{
  "patentIds": ["EP3000000B1"],
  "patentIdsMode": "citations",
  "citationDirection": "both",
  "citationDepth": 2,
  "maxResults": 50
}

📤 Output

One row per patent. Search rows carry the search-result fields; full records add everything from the patent page.

A full record (lists shortened):

{
  "patentId": "US10000000B2",
  "title": "Coherent LADAR using intra-pixel quadrature detection",
  "abstract": "A frequency modulated (coherent) laser detection and ranging system includes a read-out integrated circuit formed with a two-dimensional array of detector elements…",
  "priorityDate": "2015-03-10",
  "filingDate": "2015-03-10",
  "publicationDate": "2018-06-19",
  "grantDate": "2018-06-19",
  "expirationDate": "2036-05-13",
  "countryCode": "US",
  "kindCode": "B2",
  "type": "PATENT",
  "status": "GRANT",
  "legalStatus": "Active",
  "applicationNumber": "US14/643,719",
  "assignee": "Raytheon Co",
  "inventors": ["Joseph Marron"],
  "claimsCount": 20,
  "independentClaims": ["1. A laser detection and ranging (LADAR) system, comprising: a two-dimensional array of detector elements…"],
  "cpc": ["G01S7/4863", "G01S13/89", "G01S7/4865"],
  "citationsBackwardTotal": 18,
  "citationsForwardTotal": 57,
  "citedBy": [
    {"patentId": "US20180156897A1", "priorityDate": "2015-07-31", "publicationDate": "2018-06-07", "assignee": "SZ DJI Technology Co., Ltd.", "title": "Detection apparatus, detection system, detection method, and movable device", "examinerCited": true}
  ],
  "familyId": "55456961",
  "familyMembers": ["EP3268771B1", "JP6570658B2"],
  "familyMembersTotal": 14,
  "legalEvents": [{"date": "2015-03-10", "code": "AS", "title": "Assignment"}],
  "litigationCount": 5,
  "pdfUrl": "https://patentimages.storage.googleapis.com/c0/d5/f7/86ad5b42759506/US10000000.pdf",
  "googlePatentsUrl": "https://patents.google.com/patent/US10000000B2/en",
  "mode": "details",
  "detailed": true
}

A search row:

{
  "patentId": "US11835958B2",
  "title": "Predictive motion planning system and method",
  "snippet": "Methods and systems for training a motion planner for an autonomous vehicle are described…",
  "priorityDate": "2020-07-28",
  "publicationDate": "2023-12-05",
  "grantDate": "2023-12-05",
  "countryCode": "US",
  "kindCode": "B2",
  "status": "GRANT",
  "assignee": "Huawei Technologies Co., Ltd.",
  "inventors": ["Kasra Rezaee"],
  "pdfUrl": "https://patentimages.storage.googleapis.com/89/e0/b6/ca12f27e0adae3/US11835958.pdf",
  "googlePatentsUrl": "https://patents.google.com/patent/US11835958B2/en",
  "searchInput": "machine learning autonomous vehicle",
  "position": 4,
  "totalResults": 154527,
  "detailed": false
}

The run's key-value store also holds an OUTPUT summary: totals per search, patent numbers that were not found, and notes.

💰 Pricing

• $1.50 per 1,000 search results (patent event, $0.0015 each).
• $4.00 per 1,000 full records (patent-details event, $0.004 each). A full record replaces the search-result price for that row; it is not charged on top.
• Platform usage (compute and proxy traffic) is added on top and is small: in our tests 10,000 search results used about $0.012 and 1,000 full records about $0.06.
• Examples: 1,000 search results ≈ $1.50, 1,000 full records ≈ $4.06 including usage.
• Set a maximum charge per run and the scraper stops cleanly when it is reached; you are never charged for rows that were not saved.

❓ FAQ

Is it legal to scrape Google Patents? Patent documents are public records published by patent offices. The scraper reads only what patents.google.com shows to any visitor. Check the terms that apply to your use, especially for redistribution.

Do I need a Google account, API key or my own proxies? No. The default datacenter proxies work; requests Google rate-limits are retried automatically, falling back to US residential exits when needed.

How many results can I get? As many as the search has. Google Patents shows 1,000 results per search, so bigger searches are split into date ranges (newest first) and merged without duplicates.

Why is a field empty? Patent offices publish different data. Many EP and JP grants have no abstract on Google Patents, older patents have no legal events, and search rows list only the first inventor (turn on full records for all inventors).

What does "original language" do? English (the default) gives Google's English text, machine-translated for non-English patents. Original keeps the text in the language it was published in.

Can I schedule it or use the API? Yes. Schedule runs in the console or start them through the API with the same JSON input, then read the dataset as JSON, CSV or Excel.

⚠️ Limits

• Search rows show only the data in Google Patents' result list (first inventor, snippet instead of abstract); full records cost more because each needs the patent page.
• Above 1,000 results per search, results come in date order (newest first), not by relevance.
• The full description text is optional and can be very long.
• Patent numbers without a kind code (e.g. US10000000) are looked up through search; an unknown number is listed in OUTPUT and not charged.
• Non-patent literature, PDFs and drawings are returned as links or citation text, not downloaded.