PDF — 20 Free Online Tools
Free, browser-based pdf tools on Toolpile. No signup, no file uploads — everything runs locally on your device.
PDF is the lingua franca of documents — universal, but built for printing, not editing. Every PDF tool here picks one specific job and does it well in your browser, so confidential contracts, scanned IDs, and signed forms never leave your device. Pick the right one for what you actually need to do.
When to reach for which PDF tool
Restructuring an existing PDF (split, merge, reorder, rotate, delete pages, add page numbers): use Merge PDF, Split PDF, Reorder Pages, Rotate PDF, Delete Pages, or Page Numbers PDF. These are the most-used tools in the category because most PDF problems are layout problems — combining two scans, dropping a blank page, fixing a sideways spreadsheet from a bad export. Every one runs entirely client-side via pdf-lib; no upload needed.
Annotating (watermark, metadata): Watermark PDF stamps text across each page; PDF Metadata edits the title/author/keywords fields that show up in PDF readers and search engines. Metadata edits don't change the visual content — useful when a contract was exported with a temp filename you don't want clients to see.
Converting between formats: PDF to Image (raster the pages to PNG/JPG), PDF to Text (pull the raw text layer, OCR-free), PDF to Word (best-effort .docx export — preserves text and basic formatting; complex layouts may need manual cleanup), Image to PDF (combine scans into one document), and Text to PDF (paste plain text, get an A4 PDF). For high-fidelity .docx → .pdf conversion, see the note on the Text to PDF page.
Inspecting: Compare PDFs surfaces text differences between two versions — useful for legal redlines or contract revisions when neither party has track-changes available.
What "client-side" actually means here
Every tool in this category processes PDFs entirely in your browser using pdf-lib (or canvas for the image-conversion tools). There is no server-side processing of your files: when you click "Merge" or "Split", the work happens on your CPU, not ours. The only network traffic during a tool run is the page itself loading, plus optional analytics that never include file content. You can confirm this by opening DevTools → Network panel and watching during a conversion — you'll see no file uploads.
This matters for sensitive documents: contracts, medical scans, financial statements, anything under NDA. Most online PDF tools (SmallPDF, iLovePDF, PDF24) upload to their servers, process there, and send back. That's faster for huge files (server CPU > browser CPU) but exposes your content. The trade-off here is honest: if you have a 500 MB PDF and want it processed in 2 seconds, you'll want a server-side tool; for everything under ~200 MB total, this category processes it locally without leaking.
PDF gotchas worth knowing
Encryption blocks every tool: pdf-lib refuses to read encrypted streams. If a PDF won't open with any of these tools, it's almost always password-protected — run it through Unlock PDF first (this site has it), then use the rest of the category.
Digital signatures invalidate when content is reordered: this is by design at the PDF spec level, not a tool limitation. Any merge, split, or page-reorder operation breaks the signature. Re-sign after editing, or keep the signed document as-read-only and edit a copy.
Rotation is metadata, not actual rotation: PDF rotation is stored as a per-page flag (0/90/180/270) that viewers apply at render time. Most viewers respect it; some print drivers and OCR tools don't. If your output looks sideways downstream, run it through Rotate PDF and pick "flatten" to bake the rotation into the content stream.
Font subsetting affects merge size: when two PDFs embed the same font but at different subsets (different glyph ranges), the merged file carries both subsets. Output may be larger than expected — that's deliberate (otherwise glyphs would render as missing boxes). Use a compress step after merging if size matters.
Common workflow recipes
"I scanned a contract in two passes (front, back) and need it interleaved": Split each scan to one-page-per-file (Split PDF), drag pages into the right order in Merge PDF, output the combined version. Future cycles: a one-step "interleave" feature would replace this; for now this two-step works.
"My exported PDF has a sideways spreadsheet on page 3": Open Rotate PDF, target page 3, rotate -90°, save. Bonus: if page 3 only needs to look right in a viewer, the rotation flag is enough; if you'll OCR it or print, flatten the rotation.
"I need to send a contract but the PDF metadata still shows the temporary filename": PDF Metadata → clear or edit the Title/Author/Subject fields. Most readers display these in their tab title, so this is an under-known privacy step.
"Pull the table out of a quarterly report PDF": PDF to CSV (PDF to Excel tool) — works on PDFs whose tables are real text; if the report is a scan (image-only PDF), PDF to Text returns nothing usable and you'd need OCR upstream.