Documents change.
Stay on the same page.

Paste the document link. Review the text. Go.

1. Open Add monitor and paste the PDF’s direct public URL. A link ending in .pdf is recognized automatically. Download links are also recognized when their response identifies a PDF.
2. Review Document text. The whole document is selected automatically. The page count and readable-page count help you check what PageFluxer can actually read.
3. Choose any text change, or a phrase condition. Set a name or schedule if you need one, then start monitoring.

Your confirmed preview becomes the initial baseline. It does not send an alert for content already present. Email schedules, quiet hours and independent webhook delivery stay on the monitor details page. Optional AI summaries work with the same consent and plan allowances as webpage text changes.

Use the PDF’s own URL, not a browser’s internal viewer URL or a webpage that embeds a document. You can usually right-click the document link and copy its link address. Local files and uploads are not supported.

Choose the update worth hearing about

• The text changes: catch additions, removals and wording changes in the extracted text.
• It starts containing a phrase: get an alert when text such as “applications open” becomes present.
• It stops containing a phrase: know when a condition such as “temporary closure” disappears.

Phrase matching is case-insensitive and literal. It fires when the condition becomes true, rather than repeating while it remains true. Up to three whole-document text rules can share one check. Page regions, individual page selection, PDF price parsing and new-item extraction are not part of this version.

Text changes are retained in history even if no alert rule matches. A layout-only change or a new file timestamp does not by itself trigger an alert. PDF text order, wrapping and font encoding can still change how extracted text is compared; check the preview for complex tables and multi-column documents.

A useful free monitor. Room to grow.

A PDF uses one of your normal monitor slots. Each scheduled PDF attempt, including a failed document check, uses one HTML check from the shared allowance. It uses no Browser check credits and does not start or extend the Free browser trial. Checks pause at the monthly allowance; there are no automatic overage charges. Preview credits, history storage, email limits and plan retention still apply.

Free’s PDF access is ongoing, not a timed trial. You can pause one PDF monitor and activate another within your total slots. On a downgrade, excess active PDFs are paused, keeping the oldest eligible monitors active. Your retained history remains subject to the new plan’s limits. Service capacity and failure backoff may delay scheduled checks.

What this version can read

• A direct public HTTP or HTTPS PDF, accessible without a login, cookies or password.
• Up to 5 MiB downloaded, 100 pages, and 512 KiB of extracted text. Multiple text rules also share the existing selected-value storage limit.
• A readable text layer. Scanned or image-only documents need OCR, which is not included. Pages with no extractable text are identified in the preview and are not monitored.
• A complete capture within the document processor’s time limit. Documents are rejected rather than truncated to fit.

Monitoring compares document text. It does not compare screenshots, images, signatures, embedded attachments, interactive form state or page layout. Protected documents and unusual font encodings may be unsupported. Page loading waits, scrolling and CSS selectors apply to webpages, not PDFs.

PageFluxer does not store the original PDF file in change history. It retains compressed extracted text for comparisons under your plan’s history rules. The preview opens the original publisher’s link when you choose Open PDF. Monitoring never inherits the login or cookies from your browser or extension.

If a document cannot be captured

• No readable text: try selecting and copying a sentence from the PDF. If it is a scan, look for an accessible text version.
• Encrypted or protected: choose a public, unprotected copy. Password submission is not supported.
• Wrong file or sign-in page: copy the direct document link and check that it works without signing in.
• Too large or too slow: choose a shorter document. Selecting a phrase does not reduce the amount that must be downloaded and parsed.
• The PDF link changed: this monitor follows the saved URL, including its normal redirects. It does not search for a replacement document. Monitor the publisher’s webpage or feed for new links, then create a PDF monitor for the new URL.

A failed download, missing text layer, invalid document or processing limit preserves your last successful baseline. It does not become an “everything was deleted” change. Fix the source problem and retry the preview; scheduled failures retry with backoff.

The same document flow through the Pro API

Use POST /api/v1/previews with the document URL. Set source_type: "pdf" for an explicit PDF request, including download links without a file extension. The API keeps load_mode: "html" for shared usage accounting; the response identifies source_type: "pdf" and includes document metadata.

{
  "url": "https://example.com/policy.pdf",
  "load_mode": "html",
  "source_type": "pdf"
}

Evaluate a whole-document text target, then create the monitor with the returned preview ID using the normal API flow. The saved source type comes from the verified preview and cannot be supplied independently to bypass validation.

{
  "id": "document",
  "label": "Document text",
  "kind": "text",
  "selection": {
    "scope": "page"
  },
  "rule": "changed"
}