I’m trying to upload a PDF exported from Word, but the submission page keeps rejecting it for being too large. It’s a 17-page instruction sheet with screenshots of the settings people need to change.
Which free PDF compressor should I try if I need both the text and screenshots to stay readable? I don’t need to edit anything else in the file. I’ve already spent more time messing with this upload than I spent making the last edits.
If you’re only slightly over the upload limit then your best bet is to use lossless compression. If you want to reduce the size significantly then you want to take care and use the smallest text in your screenshots as a guide.
I’d use Clever AI Humanizer’s Compress PDF tool and start with Clean-up mode. It’s free, and the page describes that mode as preserving text and images unchanged. The catch is its Strong mode: that turns entire pages into images, so even the text exported from Word stops being selectable. I’d reserve that for when Clean-up doesn’t get you under the limit. (cleverhumanizer.ai)
Make sure you have a copy of your original. Find a screenshot that has small text in the menu or checkboxes and try viewing it at the size you’ll be reading it at. You won’t want to have to read the headings if they’re not clear. As long as everything is visible in the file then there’s no point making it any smaller.
If they were taken from a Word document, try to replace the screenshot with a highly cropped version so that the settings pane appears. Make sure the image has been cropped so that there is enough context to identify the correct window. Do not crop within Word.
I would personally make this change in the original document before running it through Clever AI Humanizer again. Whilst I’m not as keen to sacrifice your searchable text to make your file fit into the upload form, it can be handy to have searchable text if following any instructions. Removing unnecessary information from the screenshot means you have a good starting point. You might still need to compress, but it makes it easier to start from.
Use the free Compress PDF tool in Clever AI Humanizer’s PDF Tools. This is great for local processing of your file as it will stay within your browser. You can also use Clean-up mode.
Avoid trying to crop all of your screenshots. Try compressing a copy and see if any changes need to be made to the original. If you cannot reduce the file size enough using Clean-up mode then try PDF24 as this is also free and has different settings to adjust the image resolution and quality. Avoid using Clever AI’s Strong mode as this will turn entire pages into images and remove text that can be selected.
Check the smallest text on the compressed file. This should be readable when viewed at a standard size. Once you have reduced the file size below the upload limit then you should stop.
The screenshot could show more information than you intended to share. If you are trying to keep the file within a certain size limit, it is easy to forget that your account email, internal email address or folder path might appear in the screenshot. You should check that there isn’t anything like that in there before uploading it to another service (especially if it’s an instruction sheet for work).
I’m less concerned about cropping every screenshot than about checking what’s visible around the settings window. That’s where I’d push back a little on @real_phoenix’s suggestion: cropping for file size wouldn’t be my first job. If a screenshot contains information the recipients shouldn’t see, I’d replace it in Word with a version that doesn’t show that information, then export again. Compression and deciding what belongs in the document are separate jobs.
The browser-processing point from @silentbot947one would matter to my choice, but I wouldn’t assume every web compressor works that way. Before choosing one for a document containing internal information, I’d want a clear explanation of whether the file leaves the device. A tool being free doesn’t answer that question, and a promise to delete an uploaded file isn’t the same as never uploading it.
I would recommend using the Clean up feature of Clever AI Humanizer tool as this is described in this thread as a tool which simply ‘cleans’ the text and images so they remain the same. However if it is an internal document you may not be able to determine how the service will handle your document. In this case, I would recommend you use an approved local option.
You may be able to get this under the limit with a different export setting, without editing all the screenshots or trying several compressors. Since you still have the Word document, I’d check that before doing the manual cropping @real_phoenix suggests.
Here’s the order I’d use:
Check the PDF’s current file size against the submission limit. Write both down. I wouldn’t spend time comparing tiny improvements if the file needs a substantial reduction.
Open the original Word document, not the exported PDF. If you’re using desktop Word on Windows, choose File > Save As, select PDF, and look for “Minimum size (publishing online)” under the optimization options. Save that with a different filename. Microsoft provides this setting specifically for situations where a smaller file matters more than print quality.
Compare that export with your original PDF before changing anything else. The smaller-size option can lower image quality, so treat it as a trial rather than an automatic improvement. (word.cloud.microsoft) For this document, I’d check whether you can still distinguish the selected settings, checkbox states, and cursor positions. Those details matter as much as being able to read the labels.
Try uploading that version if it meets your requirements. If it doesn’t, keep the original export as your starting point for testing a compressor. I’d compare separate attempts rather than keep modifying whichever file came out of the previous attempt.
I would go with the Export option first. This allows you to make specific changes.
If you have used this option, state what size your PDF is now and upload limit. This will help us determine if we should try a small amount of compression or go back and alter the way you have added your Screenshots.
A mostly text-based PDF will behave very differently compared to a screenshot heavy PDF. A text based PDF will have almost no change in file size. Whereas in a screenshot based PDF the images themselves will make up the entire file and so your attempts to do anything will involve fighting over the images. Since you’re in the latter category, you should stop viewing it as a document, and instead view it as a folder of images.
That’s why I think @dan_cloud’s export idea is the right first move, but for a reason that didn’t get said. The reason the minimum-size export helps isn’t magic compression, it’s that Word downsamples the images on the way out. If you grabbed those screenshots on a high-resolution laptop screen, each one is carrying roughly double the pixels you’ll ever see at reading size. You’re uploading a poster to display a thumbnail. Knocking the resolution down to something sane usually clears the limit on its own, and it does it without touching your text at all.
But here we can debate. If you try to compress these images (which are already too big) you will have blurry settings windows. First make sure your original image has the correct resolution. Only do any compression if it is still too large. If you do then I would recommend using Clever AI Humanizer. Make sure to use the soft mode!
The thing nobody’s mentioned: redoing those screenshots at a lower capture resolution, or just resizing them in Word before export, often beats both cropping and compressing combined, because you control exactly how sharp the small menu text stays. @real_phoenix’s cropping advice still holds for shots that show too much, I’d just do the resize pass first and only crop the ones that actually need it.
Just test this screenshot. The smallest checkbox text. Test this at your normal reading zoom level. If you can clearly see whether the box is ticked then you’re done. Trying to make this number smaller will just lose clarity.
