ImagePDF.Tools
Technical

How to Reduce PDF File Size While Keeping Vector Text 100% Sharp (Smart Compression)

N
NikolaLast updated on October 16, 2026 · 6 min read

Summary

Why some PDF compressors turn text into blurry pictures, where a PDF's size really comes from, and how to shrink only the images in Acrobat Pro, Mac Preview, Ghostscript or your browser.

Text in a PDF is stored as characters plus an embedded font, so it has no resolution and cannot get blurry on its own. If a compressed PDF has fuzzy text, the tool turned every page into a picture. To make a PDF smaller and keep the text sharp, shrink the images inside it and leave the text, fonts and vector drawings exactly as they are.

That approach works because nearly every oversized PDF is oversized because of images: a 24 megapixel photo placed three inches wide, a pasted screenshot, a scanned page. Downsample those to the resolution they are actually shown at, recompress them, and the file shrinks while the text stays selectable, searchable and crisp at any zoom.

Title card reading: Reduce PDF file size while keeping vector text sharp | ImagePDF.Tools

How to tell whether a compressor flattened your pages

Check the compressed file before you send it. Any one of these tests gives it away:

  • ●Search it. Press Ctrl+F (Cmd+F on a Mac) and look for a word you can see. A flattened page finds nothing, unless the tool ran OCR to add a hidden text layer.
  • ●Select a line. Real text highlights word by word. A flattened page selects nothing, or selects the whole page as one image.
  • ●Zoom to 400%. Vector text stays clean. Rasterized text shows stair-stepped edges or the faint blotches JPEG leaves around dark lettering.
  • ●List the fonts. In Acrobat Reader, open File > Document Properties and choose the Fonts tab. A document that had text but now lists no fonts has been converted to images.

Where the bytes in a PDF actually are

Text is cheap. A long report with no pictures is often well under a megabyte, because the text is stored as compressed instructions and each font is usually embedded as a subset containing only the characters used.

Photos are expensive. A 6000 by 4000 pixel camera image holds 24 million pixels and commonly takes 5 to 10 MB even as a good JPEG. Placed three inches wide on a page, it only needs 450 pixels across to look sharp on screen at 150 pixels per inch. Scaling it to 450 by 300 leaves about 1/178 of the original pixels, which is where the dramatic savings come from.

A few other things bloat PDFs and are worth knowing about:

  • ●Fully embedded fonts instead of subsets. Fonts for Chinese, Japanese or Korean can be several megabytes each.
  • ●The same image stored many times, often a logo on every page after files were merged.
  • ●Editing data from design apps. Illustrator's "Preserve Illustrator Editing Capabilities" option stores a full copy of the artwork file inside the PDF.
  • ●Very complex vector art, such as detailed maps or CAD drawings with millions of path points. This is the one case where the text-safe approach cannot shrink much, because the size sits in vector content you want to keep.

Acrobat Pro can show you the breakdown: in the PDF Optimizer dialog described below, click Audit space usage to see how many bytes go to images, fonts, content streams and other objects.

Choosing a target image resolution

Resolution matters far more than JPEG quality, because file size follows pixel count. Halving the resolution leaves a quarter of the pixels. Going from 300 to 100 pixels per inch leaves one ninth.

Where the PDF is goingImage resolutionWhat to expect
Commercial printing300 ppiDo not go lower. Keep this as your master copy.
Office printer and screen150 to 200 ppiPhotos print cleanly on a laser or inkjet.
Email, web, on-screen reading150 ppiSharp on most screens. Ghostscript's /ebook preset uses this value.
Smallest possible, screen only72 to 96 ppiPhotos look soft when zoomed. Text is unaffected.
Pixels per inch measured at the size the image is printed on the page.

Method 1: PDF Optimizer in Acrobat Pro

  1. 1.Open the PDF and choose Menu > Save as Other > Optimized PDF. In older versions it is File > Save As Other > Optimized PDF.
  2. 2.Click Audit space usage to confirm images are the problem before you change anything.
  3. 3.In the Images panel, set color and grayscale images to Bicubic Downsampling to 150 pixels per inch for images above 225, with JPEG compression at Medium or High quality.
  4. 4.In the Fonts panel, leave fonts embedded. Unembedding saves a little space and risks the text changing on someone else's computer.
  5. 5.In Discard User Data, tick the option to discard private data of other applications. This removes the Illustrator and Photoshop editing data mentioned above.
  6. 6.Click OK and save under a new name, so the original stays untouched.

Acrobat's simpler Compress a PDF tool also leaves text as text. The optimizer is worth the extra clicks when you need a specific resolution or want to know what changed.

Method 2: Preview on a Mac

Open the PDF in Preview and choose File > Export. The Quartz Filter menu offers Reduce File Size, which recompresses images and keeps text as text. Apple warns the result may be lower quality than the original, and in practice this filter is hard on photos, which can come out blocky. If that happens, export again without the filter and tick Optimize images for screen and Save images as JPEG instead.

Method 3: Ghostscript, free on any operating system

Ghostscript rewrites a PDF with its own settings. Text stays text, and images are downsampled according to a preset:

bash
gs -sDEVICE=pdfwrite -dPDFSETTINGS=/ebook -dNOPAUSE -dBATCH -dQUIET -sOutputFile=smaller.pdf input.pdf

/ebook downsamples color and grayscale images to 150 dpi, /screen to 72 dpi, and /printer leaves them at 300 dpi. For a value in between, replace the preset with -dDownsampleColorImages=true -dColorImageResolution=200 and the matching Gray options. Because Ghostscript builds a new file rather than editing the old one, open the result and check form fields, links and bookmarks before you rely on it.

Method 4: In your browser with ImagePDF.Tools

Compress PDF works the same way as the optimizer: it re-encodes the embedded images and leaves the text, fonts, links and form fields alone. For each image it works out the size it is drawn at on the page, resamples it to the level you pick (Light is 180 DPI, Medium 150 DPI, Heavy 96 DPI) and keeps the new version only if it is actually smaller. It also stores repeated images once, recompresses the other internal streams and removes document metadata. If you have a hard limit, the Target file size option works backwards to it.

After compressing, the tool reopens its own output and compares the page count and text with your original. If they do not match, you get the original back instead of a damaged file. Everything runs on your device, so the PDF is never uploaded. Password-protected files need to go through Unlock PDF first.

💡

If a text-only PDF barely shrinks, nothing is wrong. There were no large images to work on, and the remaining bytes are fonts and text, which you want to keep.

Fix the source file when you can

Exporting a lighter PDF in the first place beats compressing it afterward. In Word for Windows, choose File > Save As, pick PDF and select Minimum size (publishing online), or use Picture Format > Compress Pictures before exporting. In InDesign, the Smallest File Size preset downsamples images to screen resolution. In Illustrator, untick Preserve Illustrator Editing Capabilities when you save a PDF for sharing. Scanned documents are a different case, since every page is already an image; our guide to oversized scanned PDFs covers them.

Frequently asked questions

Will compressing remove the OCR text layer from a scanned PDF?
Not with a compressor that only re-encodes images. The OCR text is stored separately from the page image, so it survives. Tools that rasterize whole pages do remove it, because they replace everything on the page with a new picture.
Why is my compressed PDF bigger than the original?
Some tools re-encode every image even when the original encoding was already efficient, or rasterize vector pages into large images. A careful compressor compares each new image with the old one and keeps whichever is smaller.
Can I shrink a CAD drawing or map PDF without rasterizing it?
Only modestly. Recompressing streams and saving with object streams trims some overhead, but the bulk is path data you want to keep. Real savings mean simplifying the drawing in the source program, for example by removing hidden layers or reducing curve detail.
Is it safe to compress the only copy of a document?
Downsampling throws image detail away for good, so keep the original. Save the compressed version under a new name and use it for email or upload, and keep the full-resolution file for printing or later edits.

Sources & references

This article was researched and written by Nikola, drawing on the following primary sources and documentation:

Ready to try it?

All tools run entirely in your browser, no uploads, no account required.

Compress PDF