Google Patents Scraper: Titles, Abstracts, Claims & Assignees

Turn any Google Patents search into a clean, structured dataset of patents. Search by keyword and narrow by inventor, assignee or priority-date range, and get title, abstract, claims, inventors, assignee, filing/publication dates and CPC codes for every result, fully paginated.

Perfect for patent landscaping, competitor IP monitoring, prior-art search, and technology scouting across any field.

What you get

• 🔎 Search by keyword, narrowed by inventor, assignee, or priority-date range.
• 📄 Full record per patent: title, abstract, and every claim.
• 🏢 Inventors and assignee (the patent owner), in canonical English.
• 🗓️ Priority, filing, publication and grant dates, plus application number.
• 🏷️ CPC classification codes and issuing-authority country code.
• 📊 Clean JSON / CSV / Excel export, ready for a spreadsheet or a pipeline.

Input

Field	Type	Description
searchQueries	string[]	Keyword searches to run, e.g. self-driving vehicle. Required.
inventor	string	Restrict every query to this inventor name.
assignee	string	Restrict every query to this assignee (patent owner), e.g. Waymo.
dateFrom / dateTo	string	Priority-date range, YYYY-MM-DD.
maxItems	int	Stop after this many patents across all queries (default 100).
proxyConfiguration	proxy	Datacenter rotation by default; switch to residential only if rate-limited.

Example input

{
  "searchQueries": ["autonomous vehicle"],
  "assignee": "Waymo",
  "dateFrom": "2020-01-01",
  "maxItems": 100
}

Output

One record per patent, deduplicated by publication number.

Field	Description
publicationNumber, title, url	Patent identity and its Google Patents page.
abstract, claims	Abstract text and one entry per claim.
inventors, assignees	Inventor name(s) and original assignee(s).
priorityDate, filingDate, publicationDate, grantDate	Key dates.
applicationNumber, countryCode, cpcCodes	Application number, authority, and CPC codes.
pdfUrl, snippet	PDF link and the search snippet that matched.
query, scrapedAt	Source query and scrape time.

Example output

{
  "publicationNumber": "US12525077B2",
  "title": "Self-driving vehicles and weigh station operation",
  "url": "https://patents.google.com/patent/US12525077B2/en",
  "abstract": "The technology involves operation of a self-driving truck...",
  "inventors": ["Vijaysai Patnaik"],
  "assignees": ["Waymo LLC"],
  "priorityDate": "2019-12-16",
  "filingDate": "2023-11-14",
  "publicationDate": "2026-01-13",
  "cpcCodes": ["B60W50/0205"],
  "query": "self-driving vehicle"
}

FAQ

Do I need a Google account or API key? No. Just provide a search query.

How many results per query? Google Patents pages results in tens and caps a single query at 1000 results; the actor paginates up to your maxItems cap.

Why are some abstracts or claims empty? Patents from some non-US offices don't expose an English abstract or machine-readable claims on Google Patents; core bibliographic fields are always populated.