Compress a large scanned PDF
A scanned book is heavy because every page is a grey photo. Turn the text pages into clean black and white and the file shrinks without losing sharpness, because black type on white paper needs one bit a pixel instead of twenty-four.
Key takeaways
- Most compress buttons lower image quality until the size drops. On a page of type that shows up as fuzzy letters.
- FixMyScan takes files up to 1 GB. Free tiers elsewhere stop sooner: Sejda at 50 MB, iLovePDF at 300 MB.
- Measured on 7 Nov 2026: a 500-page colour scan went from 70 MB to 27 MB, 62% smaller, with OCR on every page.
- Picture pages stay greyscale, so plates and maps survive. A book full of photographs shrinks less.
- The smaller file is searchable as well: OCR runs on every page.
Where the megabytes go
Open a 300 MB scanned book and almost none of that weight is words. It is paper: the cream tint of every page, the show-through from the back of the leaf, the shadow in the gutter, all recorded as millions of grey or colour pixels. Library and archive scans are made at high resolution for keeping, so a long book easily passes the size an upload box or an email will take.
The type itself is simple. Each pixel is either ink or not. Store it that way and a page shrinks to a fraction of its weight.
Two ways to shrink a scan
| Squeeze the photos | Clean to black and white | |
|---|---|---|
| What changes | Lower JPEG quality, lower resolution | Paper levelled to white, print to solid black |
| Text | Softer, with smudges round letters | Crisp black on white |
| Size | Smaller | 62% smaller in our test below |
| Pictures | Squeezed too | Kept in greyscale |
| Search | Unchanged | OCR added on every page |
The first is what a general compress tool does, because it can’t tell a page of text from a photograph. The second only works if the tool knows it is looking at a book.
A measured example
On 7 Nov 2026 we ran a real scan through FixMyScan’s clean-up: Ernest Thompson Seton’s Two Little Savages, from archive.org, 500 pages saved as colour photos at 150 dpi, with the text layer taken out.
| Before | After | |
|---|---|---|
| File size | 70.5 MB | 26.8 MB |
| Text pages | Colour photos | 479 pages in sharp black and white |
| Picture pages | Colour photos | 21 pages kept in greyscale |
| Search | None | OCR on every page |
That is 62% smaller, and under the 50 MB limit for sending to a Kindle by email. A scan made at a higher resolution has more weight to lose. A book that is mostly plates shrinks less.
Large files and upload limits
Size is often the first wall a long book hits. Here is what each tool states (Sejda and iLovePDF checked 16 Oct 2026, Gemini and Claude 7 Nov 2026):
| Tool | Largest file |
|---|---|
| Sejda, free | 50 MB |
| iLovePDF, free compress | 300 MB |
| Gemini app | 100 MB |
| Claude | 500 MB, PDFs to 1,000 pages |
| FixMyScan | 1 GB, books to 1,000 pages |
How to compress a scanned book with FixMyScan
- Upload the PDF or DjVu. Up to 1 GB. The file goes straight to private storage, not through a web form.
- Set the crop once. One box trims the margins of every page, which also removes the dark edges that add weight.
- Let it run. Text pages are cleaned to black and white and read with OCR. Pages with pictures stay greyscale.
- Check the marked pages. Swap any page back to the original if you prefer it.
- Download the smaller PDF, or send it to your Kindle, Kobo or reMarkable.
When it won’t shrink much
- The scan is already black and white. A 1-bit archive scan has little left to lose. You still gain OCR and cleaner pages.
- The book is mostly pictures. Photo pages stay greyscale on purpose.
- The PDF isn’t a scan. A PDF made from a word processor is already small; a general compress tool suits it better.
Questions
How do I compress a large PDF without losing quality?
For a scanned book, don't squeeze the photos harder. Clean the text pages to black and white instead: each page gets sharper and smaller at once, and FixMyScan stores those pages without lossy compression.
What is the largest file I can upload?
1 GB, and books up to 1,000 pages. PDF and DjVu both work.
Will the pages still look like the book?
Yes. Pages keep their layout, order and page numbers. The paper turns white and the print black, and any page can be swapped back to the original if you prefer it.
Why is my scanned PDF so big?
Each page is saved as a photo, usually in colour or greyscale and at high resolution. A 500-page book of plain text can weigh as much as a holiday's worth of pictures.
Can ChatGPT or Claude compress it for me?
Only if you can attach it. Gemini's app takes files up to 100 MB and Claude up to 500 MB. We tested what each one does with a scanned book in this comparison.
Is it small enough for my Kindle?
Send to Kindle takes 200 MB on its website and 50 MB by email. Our will it fit check compares your file with each route.
What happens to my file?
It stays private to your account. Every upload is deleted 24 hours after the book is ready, or 24 hours after upload if you don't convert it.