Add your CV, cover letter and certificates in the order a recruiter should read them, state the portal's size cap, and the chain merges them, compresses toward the cap while keeping text selectable, then re-opens the file it produced to check the page count, the size and the text layer. Nothing is uploaded.
UnboundPDF is a free suite of 43 PDF and image tools that run entirely in your browser — merge, split, compress, edit text, OCR in 126 languages, redact, sign, convert and archive to PDF/A. Your document is read and written by the page on your own device; there is no document-upload endpoint in the core tools, no account, no watermark and no daily cap. Every result can be checked — with the Network tab, or with the Document Passport the Workspace writes for a chain of steps.
How it works, step by step
- Add your CV, cover letter and any certificates, in reading order.
- State the size cap the portal enforces, in MB.
- Run the chain — it merges, then compresses only as far as it must.
- Read the receipt: page count, size against the cap, and whether text survived.
- Download the pack, and the Document Passport if you want a record.
What the chain actually does
Three things, in one pass, each of them an engine that already powers a tool here. First Merge PDF copies every page of every file you added into one document, in the order shown in the list. Then Compress PDF’s Balanced engine re-compresses the photos the merged file embeds — re-quantising them in the JPEG coefficient domain rather than decoding and re-encoding, so text, fonts and vector graphics are never touched. Finally the produced bytes are re-opened and measured. Nothing is assumed: the page count comes from re-parsing the output, not from adding up what went in.
Why the cap is the interesting part
Portals often state a cap and reject silently above it. The chain escalates only as far as it must: it works out the reduction the cap implies, tries that, tries harder if the file is still over, and stops the moment it is under. If it never gets under, the receipt says so with both numbers rather than handing you a file that will bounce. Tighter targets have their own walkthroughs — 2 MB and 1 MB — and the rules that government and visa portals apply are collected in compressing for government portals.
Text has to survive, or the parser sees nothing
An applicant tracking system reads the text layer, which is why re-rendering pages as images is off by default here: it wins on size and loses the job. The receipt reports how many characters of real text the output still carries, so “text survived” is a number rather than a promise — and if you do allow image mode, the same line will show the text gone before you submit.
The verification receipt
After the chain finishes, five checks run against the produced file: it re-opens in the same reader this site renders with; its page count matches the sum of the inputs; its size is compared with the cap you stated; its encryption state is read (nothing here encrypts, so anything but “not encrypted” is a bug worth catching); and its text layer is sampled. Failures are printed as failures with the real numbers. A check that could not run says “not checked” — never a quiet pass.
The Document Passport, and what it proves
A small JSON file: the chain that ran, every check with its real result, and the SHA-256 of the file you downloaded. That hash proves exactly one thing — these bytes have not changed since. It does not prove who wrote the document, that anyone approved it, or that anything in it is true. It is self-issued and unsigned, and we would rather say so than let a hash look like a certificate.
When to do it by hand instead
If you need to reorder pages, drop a page, or rotate a scan that came in sideways, the chain is the wrong shape — use Organize PDF first and bring the result back here. If the certificate is a photo rather than a PDF, JPG to PDF converts it first. If a portal wants each document separately, skip the merge entirely and use compress on each file. And if the CV itself needs a wording change, that is editing PDF text, which keeps the document’s own fonts.
Nothing leaves your device
A job application is a dossier: your address, your history, sometimes your passport scan. Every step of this chain runs in the browser tab, which is exactly why it can be used on documents you would not post to an upload form. The reasoning behind that model is in how we compare with upload-based tools, the verification method is in the PDF Privacy Lab, and the same private route done tool-by-tool is described in merging PDFs privately. Students assembling coursework packs may prefer the walkthrough in PDF tools for students.
Frequently asked questions
Why does keeping text selectable matter for a job application?
Most application portals run the file through an applicant tracking system that reads the text. A PDF whose pages have been re-rendered as images has no text to read, so the parser sees an empty document. This chain keeps text selectable by default and only falls back to page images if you explicitly allow it.
What order do the documents end up in?
The order you added them, top to bottom, and the list is shown before you run anything. A CV first, then the cover letter, then certificates is the common reading order — but the chain does not decide that for you.
What if the pack cannot reach the size cap?
The receipt says so, with the real numbers: the size it reached and the cap it missed. It never reports a cap it did not measure. You can then allow page images, split the pack, or send fewer documents.
Is my CV uploaded anywhere?
No. Every step runs in this browser tab on your own device — UnboundPDF has no server that receives documents. You can watch the Network tab while the chain runs; the method is written up in the PDF Privacy Lab.