Download a Google Patents publication PDF

Site patents.google.comTask download-google-patent-pdfVersion v1Updated Jul 31, 2026Category document-download

Resolve and download the PDF for a Google Patents publication using its publication number, with a caller-specified output filename. This skill was captured from a live agent session on patents.google.com and publishes here verbatim, exactly as an agent receives it.

NoteSelectors and URL schemes drift as sites change. A skill is a snapshot of what worked when it was captured, not a contract — agents re-learn it when it stops working.

Purpose

Download the PDF associated with a Google Patents publication number and save it under the requested filename.

When to Use

Use when the caller provides a Google Patents publication number such as {publication-number} and wants the original patent document as a PDF. The application number or title may be used for caller-side verification, but the publication number is the stable URL key.

Workflow

  1. Navigate directly to https://patents.google.com/patent/{publication-number}/zh (use another language suffix only if requested).
  2. In the same browser session, run this evaluator on the loaded patent page to resolve the actual PDF URL:
(() => {
  const meta = document.querySelector('meta[name="citation_pdf_url"]');
  let pdfUrl = meta && meta.content ? meta.content.trim() : "";
  if (!pdfUrl) {
    const link = Array.from(document.querySelectorAll("a[href]")).find((a) => {
      const href = a.href || "";
      return (
        /patentimages|\.pdf(?:$|[?#])/i.test(href) ||
        /download\s*pdf/i.test((a.textContent || "").trim())
      );
    });
    pdfUrl = link ? link.href : "";
  }
  const publication = decodeURIComponent(
    location.pathname.split("/").filter(Boolean)[1] || "publication",
  );
  return {
    pdfUrl,
    filename: `原始申请-${publication}.pdf`,
    publicationUrl: location.href,
  };
})();
  1. Download pdfUrl using the browser's download/save facility and store it as the caller-specified filename (for example, 原始申请-{publication-number}.pdf). Do not rely on clicking the visible Download PDF text; the metadata URL is the shortest reliable resolution path.

Site-Specific Gotchas

  • Google Patents exposes the document PDF through meta[name="citation_pdf_url"], commonly pointing to a patentimages.storage.googleapis.com URL; that storage URL is not safely guessable from the publication number.
  • The publication-number URL is case-sensitive in practice for reliable matching; preserve the supplied identifier, including its country code and kind code.
  • The page may be localized, but the PDF URL is generally independent of the /zh language suffix.
  • Verify the resolved page/publication before saving when the caller also supplies an application number or title; do not substitute a search result with a merely similar title.

Expected Output

A downloaded PDF from the resolved citation_pdf_url, saved under the caller's requested filename, plus the source publication URL and resolved PDF URL if reporting metadata.

Call it

GET https://production-sfo.browserless.io/skills?token=TOKEN-HERE&domain=patents.google.com&task=download-google-patent-pdf