Logo

Convert PDF to PDF/A

Your filing was rejected because the PDF was not archival. A normal PDF can borrow fonts and colours from the computer that opens it, so it drifts over time. Convert it to PDF/A and everything it needs travels inside the file.

Upload your PDF

Drag and drop your document here, or click to browse. Up to 100 MB.

Processed on our server, then deleted

How our PDF to PDF/A converter rebuilds your document

We do not add a label to your file. We rebuild it. A standard PDF is allowed to point at things it does not contain — a font sitting on your laptop, a colour profile belonging to your monitor, a script that runs when the file opens. Every one of those is a dependency that can vanish.

Our engine takes the page content apart and reassembles it with those dependencies resolved. Fonts get subset and embedded. Device-dependent colour becomes device-independent colour with an sRGB profile attached. Scripts, embedded media and encryption come out entirely, because the standard forbids them.

Here is the part most converters stay quiet about. When the engine meets something it cannot legally keep, it drops that element and carries on rather than failing the job. That is usually the behaviour you want on a deadline. It also means an interactive form button or an embedded video can disappear without a word, so we say it plainly on the upload screen and we are saying it again here. Open the output next to the original before you file it anywhere that matters.

Which PDF/A conformance level should you pick?

Pick the one your recipient named. When nobody named one, pick PDF/A-1b.

The parts of the standard were published in order, and each later one relaxes the rules a little. PDF/A-1b came first and is the strictest: no transparency, no layers, no attachments. PDF/A-2b added transparency, layers and JPEG 2000 compression. PDF/A-3b is PDF/A-2b with one change, the ability to carry attached files of any type.

In our experience the vague requirement is the common case. A registry or a repository says “submit as PDF/A” and stops there. A 1b file satisfies almost any system built for a later part, so it is the safe answer — with one real cost. Because 1b forbids transparency, a document with soft shadows or a semi-transparent logo comes back visibly flattened. If your file is design-heavy and the recipient has not specified, ask before you assume. We walk through the trade-offs properly in which conformance level you need.

What makes an ISO 19005 compliant PDF different

Four things, and they all point the same direction: remove every reason the file might render differently later.

  • Fonts live inside the file. A normal PDF can name Arial and trust your system to supply it. An archival one carries the glyphs it uses. This is also the single most common reason a conversion fails — see fonts not embedded error.
  • Colour is defined, not assumed. The file carries an ICC profile saying exactly what its colours meant, so a printer in 2050 can reproduce them.
  • Nothing executes. No JavaScript, no launch actions, no embedded audio or video. Anything that runs is something that can break or attack.
  • Nothing is locked. Encryption is forbidden outright, because a password is a guaranteed way to lose a document once everyone who knew it has moved on.

That last rule surprises people. If your source file is password-protected, no tool on earth can turn it into a compliant archive while it stays locked. The protection has to come off first, assuming the file is yours to unlock.

Why archives insist on a self-contained PDF

Because they have been burned. Ask anyone who has run a records office for a decade and they will describe opening a twenty-year-old contract to find the text rendered in a substitute typeface, with the line breaks in different places and a signature block that no longer lines up.

Nothing was corrupted. The file simply asked for a font that the reader could not supply, and the reader guessed. On a marketing brochure that is annoying. On a contract where the page break determines which clause sits above a signature, it is a legal problem.

A self-contained file removes the guessing. Everything needed to draw the page travels with the page. If you want the longer background on how the standard came about and what it does not cover, we wrote what PDF/A actually is.

Three steps to archive your document online

You can convert PDF to PDF A online here without installing anything, without an account, and without a watermark on the output.

  1. Upload the file. Drag it into the box at the top of this page. The limit is read from the server and shown on the upload panel, so the number you see is the number that actually applies.
  2. Choose your level. Three buttons, with the trade-off written under each one. PDF/A-2b is preselected as a sensible middle ground.
  3. Convert, then check. Download the result and confirm it before you submit it. Our PDF/A validator reads the file in your browser and tells you what it found.

Step three is not padding. We would rather you catch a problem here than discover it in a rejection email. If you are preparing a thesis, the route through Word has its own pitfalls, covered in saving a Word document as PDF/A. Scans need different handling again — a converted scan is compliant and still unsearchable, which we explain in turning scanned PDFs into PDF/A.

Where the archive PDF format falls short

No tool page should read like a sales sheet, so here is the honest list of what this format costs you.

Files usually get bigger. Embedding fonts adds weight, and a document using several typefaces or a CJK font can grow a lot. If size matters, compress the PDF first and archive second. Doing it the other way round tends to undo the compression.

Editing gets awkward. Readers open PDF/A files in a protected mode that blocks changes, which people reasonably experience as the file being broken. It is not. You can turn a PDF/A back into a normal PDF when you need to work on it, then archive the finished version. The workflow is covered in editing a PDF/A file.

Compliance is not accessibility. A PDF/A file at level b says nothing about screen readers. Only level a adds a tagged structure, and that information has to come from the program that authored the document. Our engine produces level b and we do not pretend otherwise.

We do not certify anything. We run a well-established archiving engine and report what it did, including when it could not attach a colour profile. We do not run a formal audit on your file and we will not tell you it passed one. That is what a validator is for.

Your document, still readable in thirty years

Converting PDF to PDF/A buys one specific thing: a file that carries everything it needs, so no future computer has to guess. That is why courts, universities and records offices ask for it, and it is the whole reason the format exists.

Upload your file above, pick the level you were asked for, and check the result before you send it. If you have several documents to combine first, merge them into one document and archive the finished file rather than the parts.

Frequently Asked Questions

Why does my converted file still fail a strict PDF/A validator?

+
Almost always the output intent. PDF/A needs an embedded ICC colour profile that tells future readers exactly what the colours meant, and a converter that skips it produces a file that looks right and validates wrong. We attach an sRGB output intent when the server has the profile available, and our result panel tells you when it could not. If yours says the intent is missing, run the file through a validator before you submit it anywhere.

Can you produce PDF/A-1a or PDF/A-2a rather than the b levels?

+
No, and we would rather say so than let you find out after a rejection. Level "a" requires a fully tagged document — a machine-readable reading order, alternative text on images, real heading structure — and that information has to come from the authoring program. Our engine converts what is in the file; it cannot invent a structure tree that was never there. If your requirement specifies 1a or 2a, export it correctly from Word or InDesign in the first place.

What gets removed from my document during conversion?

+
Anything the standard forbids: JavaScript, embedded audio and video, 3D objects, external file references, and encryption. Our converter drops those elements and carries on rather than failing the whole job. That is usually what you want, but it does mean a form button or an embedded video can disappear quietly, so compare the output against the original before you file it.

Why did my PDF get bigger after converting it?

+
Because self-contained files carry their own fonts. A normal PDF can reference Arial and trust your computer to supply it, while a PDF/A has to embed the glyphs it uses. Documents with several typefaces, or with a CJK font, can grow noticeably. If size matters more than speed, compress the file first and convert second — doing it the other way round can undo the compression.

My file is already PDF/A. Is converting it again harmful?

+
It is rarely useful and occasionally destructive. A second pass re-renders the page content, which can shift hairline rules and re-compress images that were already compressed once. Check what you have with a validator first. Convert again only when the level is wrong for your requirement, not as a precaution.

The requirement just says "PDF/A" with no level. Which one is safe?

+
Pick PDF/A-1b when the instruction is vague. It is the oldest and strictest part of the standard, so a file that satisfies 1b satisfies almost any system built for a later one. The trade-off is that 1b forbids transparency and layers, so a design-heavy document may come back visibly flattened. When the receiving system names a level, always use the one it names.

Does converting a scan to PDF/A make the text searchable?

+
No, and this catches people out constantly. PDF/A has never required a text layer, so a scanned page converted to PDF/A is fully compliant and still completely unsearchable. You need OCR to add that layer, and it has to happen before archiving. Run the scan through OCR first, then convert the result.

Why will you not convert my password-protected PDF?

+
The standard forbids encryption outright, because a password is a guaranteed way to lose access to a document in thirty years when nobody remembers it. There is no way to produce a compliant file that is still locked. Remove the protection first, assuming the file is yours to unlock, then convert the unprotected copy.

Do bookmarks, links and form fields survive the conversion?

+
Bookmarks and internal page links normally come through intact. External web links usually survive as annotations, though some readers stop treating them as clickable once the file is in archival mode. Interactive form fields are the real casualty: they are commonly flattened into static text, so fill the form in before you archive it, not after.

Will the converted file open in old PDF readers?

+
PDF/A-1b is based on PDF 1.4 and opens in almost anything, including readers from the mid-2000s. PDF/A-2b and 3b are based on PDF 1.7, which needs a reader from roughly 2008 onward — not a real constraint today, but worth knowing if you are submitting to an institution running old software. Any PDF/A file also opens as an ordinary PDF, so nobody needs special software to read it.

Related Tools You Might Like

Advertisement