Make a PDF ready for upload — check it, fix it, check again

Upload portals commonly turn a PDF away for a few reasons. The file is over the size limit or locked, it still has fillable fields, it carries scripts or attachments, or a scanned page has no text to read. Upload-ready checks for the ones your destination cares about: an email attachment, a job application, a government or visa portal, an exam or ID form, court e-filing, a long-term archive, or your own custom rules. Each check is marked as passing, fixed automatically or needing you. One click runs every automatic fix, then the file is checked again.

The limits are starting points, not a portal’s official rules. Set the number your portal states, and add a page-size rule if it asks for A4 or Letter. You get the fixed file and a plain-text report of every check and change, before and after. Checking, fixing and the report all happen in your browser tab. The file is never uploaded, so there is no server copy to delete.

How it works

1. Drop your PDF, open Upload-ready and pick where the file is going
2. Type the size limit the portal states and read the checklist. If a check needs you, enter the password or run OCR
3. Click Fix and check again, then download the file and its plain-text report

Pick the destination, then adjust the rules

Upload-ready starts by asking where the file is going. Each destination turns on the rules that kind of system usually enforces, plus, for every destination except Long-term archive, a starting size limit. The screen tells you that portals differ, so change the number to the one your portal states and add or remove rules. The checklist updates straight away.

The 15 MB email figure allows for the extra third that email encoding adds, keeping the message under Outlook’s 20 MB and Gmail’s 25 MB limits. The recipient’s mail server can have a lower cap, often 10 MB, so lower the number if you’re not sure — compress for email covers the common limits. The exam destination is for limits of 100–300 KB. At that size, if no colour setting fits, the last compression steps turn pictures grayscale — text stays text. Every destination also has a page-size setting (Any, A4 or Letter, A4, or Letter), which starts at Any.

What is fixed automatically, and in what order

Everything the checklist marks “fixed automatically” is handled in one run. Each step runs only when its rule is on, and most run only when the first check found a problem. PDF/A conversion runs whenever the PDF/A rule is on, and Fit to size runs whenever the file is over the limit by then, because earlier steps can add size. The order is fixed, so each step works on the output of the one before. The size is measured last, on the file as it will download.

1. Restrictions removed. A file that opens without a password but blocks printing or copying is freed. A file that needs a password to open waits until you type it.
2. Fillable fields flattened, so the answers become part of the page.
3. Pages fitted onto A4 or Letter if a page-size rule is set and some pages don’t match. Only those pages change, each keeps its orientation, content is scaled to fit and centred, and links and bookmarks stay.
4. Scanned pages recognised: pages with no text layer get one, from on-device OCR in the language you pick, laid invisibly over the page — while today’s OCR allowance covers them.
5. Scripts and launch actions removed: document JavaScript, JavaScript or Launch open actions (including ones chained after other actions), bookmark actions, additional-actions dictionaries (including form keystroke, format and calculate scripts) and links whose action is JavaScript or Launch.
6. Embedded attachments removed: the embedded-files tree, associated files and file-attachment annotations — with their data, not just the links to it.
7. Metadata removed: author, title, subject, keywords, creator, producer and any other identifying document-information entries, plus the XMP packet.
8. PDF/A-2b conversion: an sRGB output intent, XMP identification and a document ID. Only pages that draw visible text in a font that isn’t embedded are rendered at 200 DPI; the invisible OCR text layer and the free-plan line don’t cause that.
9. Fit to size, the same search Compress uses: lossless cleanup first, then the highest-quality image setting whose real output is at or under the limit.

The finished file is then checked again from the start. On the free plan, the “Made with Bindery” line is added before PDF/A conversion and the size step, so the limit holds for the file you download. The text-layer check ignores that line, so it can’t make a scanned page look searchable.

What needs you, and what the report says

A few things need you. If a file won’t open without a password, a password field appears: type the password you know and the file is unlocked on your device — Bindery can’t open a file without it. Personal identifiers the checklist finds, such as a Social Security number, are listed by page and last four digits, with “Open Redact with these marked” to box them for review; a number typed into a form field is flattened onto the page first, so it can be removed with it. Pages with no text layer, usually scans, are handled in the same pass while today’s OCR allowance covers them: the text is recognised on your device, in the language you pick, and laid invisibly over the page, in a form PDF/A accepts. The free plan includes 5 OCR pages a day and Pro has no limit; beyond the allowance, the pages are listed with a “Run OCR on the pages without text” button that reads only those pages, and you come back to the same rules.

When the run finishes, the heading says either that the file is ready for your destination or how many checks still need you, with a before-and-after size bar if you set a limit. The download keeps your file’s name with “-ready” added. Next to it is a plain-text report in three parts, Before, Changes and After, covering every check. It ends with a reminder that portals can have rules the check doesn’t cover, such as file-name patterns, page counts and colour. “Save these rules as my default” keeps the destination and its rules for next time.

Upload-ready in the browser versus upload sites

Honest limits

• It doesn’t know your portal’s exact rules. The destinations are starting points. The number that matters is the one the portal states, and you set it.
• It doesn’t rename files. If a portal wants a particular file-name pattern, rename the download yourself.
• It doesn’t guarantee acceptance. Passing every check means the file meets the rules you selected, nothing more.
• The PDF/A check is a practical preflight, not a certified validator. For a formal verdict, run the file through veraPDF. PDF to PDF/A explains the conversion in more detail.
• Some files can’t get small enough. The report records the smallest size reached. To go further, send fewer pages with Extract, or split the file if the portal accepts several.
• PDF/A can still turn some pages into images. A page that draws visible text in a font that isn’t embedded is rendered at 200 DPI so the file conforms, and its text is then no longer selectable; the report says how many.
• OCR can misread. The recognised text is invisible, so search for a name or number in the result to check it; the free plan covers 5 OCR pages a day.
• Only form fields are flattened. Comments and ordinary links stay; Flatten can remove them if you turn on annotation removal.
• With PDF/A on, its own metadata stays. Author, title and other identifying entries are removed, but PDF/A needs its XMP identification packet and records its producer.
• Identifiers are found in text, not in pictures. Scanned pages are searched once they have been recognised; a number that is only part of a photo or a signature is not.

Questions people ask

Will the portal accept my PDF after this?

Bindery can’t promise that, and doesn’t. It checks the rules you selected, fixes what can be fixed and checks again, and the report shows the result. Portals can also have rules the check doesn’t cover, such as file-name patterns, page counts and colour, so read the destination’s instructions too.

What if the file still can’t get under the size limit?

Fit to size looks for the best-quality version that fits. If none does, the file is compressed as small as it will go, the report says it is still over the limit, and the size check stays open. The next step is to change what the file contains: extract only the pages the portal asked for, or split the file if the portal accepts more than one.

Is my file uploaded to check or fix it?

No. The checklist, every fix, the second check and the report all run in your browser tab. The file never leaves your device, and there is no server copy to delete.

Why would a portal reject a PDF that opens fine on my computer?

Usually for one of the reasons Upload-ready checks: the file is over the size limit, has a password or restrictions, still has fillable fields, carries scripts or attachments, or has scanned pages with no text layer. Your PDF viewer may open it without complaint, but the upload system applies its own rules. Pick the destination and the checklist shows which of these apply to your file.

Will my photos or scans turn grey to meet a small limit?

Only if colour can’t fit. Fit to size tries the lossless cleanup first, then searches a ladder of ten settings, from 220 DPI down to 50 DPI, for the highest-quality one whose real output fits. The last three are grayscale, and one of them is chosen only when no colour setting reaches the limit. The size step never turns text into an image.

Can I check a file without changing it?

Yes. The checklist appears as soon as the file is open, and it doesn’t use one of the free plan’s 3 daily jobs. If every check passes, the button reads “Nothing to fix” — the file already meets the rules you selected. The fix run counts as one job, including any OCR it runs; the recognised pages also count toward the 5 OCR pages a day.

What does the free plan include?

Every destination and rule, with files up to 25 MB, 3 completed jobs per UTC day and 5 OCR pages a day. Exported PDFs carry a small “Made with Bindery” line at the foot of each page. It is added before PDF/A conversion and the size step, so the limit still holds, and its font is embedded, so PDF/A keeps your pages as they are. Pro removes the caps and the line.

Can it find Social Security, card or account numbers before I file?

Yes — turn on “No exposed ID or account numbers” (it is on for Court e-filing). Bindery reads the text layer on your device and lists Social Security numbers, card numbers (Luhn-checked), IBANs (checksum-checked) and full dates of birth, by page and last four digits only. “Open Redact with these marked” boxes them for you to review, and redaction then removes the text underneath. Scanned pages need OCR first, and minors’ names are not detected — check those yourself. US federal courts allow only the last four digits of Social Security and financial-account numbers and the year of birth.

Can I make a whole folder upload-ready?

Yes. In Pipelines, add the Upload-ready step — or Fit under a size — and run it over the files you drop, or point a watch folder at the recipe. With Upload-ready, each file is checked, fixed and checked again, and the ZIP carries a report for every file; Fit under a size adds a size report listing each file before and after. Outputs are named after the file they came from. Free pipelines take two files at a time; Pro removes the limit.

Guides for this job

More Bindery tools