How does size-based PDF splitting divide a PDF into parts by MB?
Size-based splitting estimates each page’s marginal weight across shared PDF streams, packs consecutive pages into candidate groups, and verifies the exact saved byte length of every output file. If shared fonts or images push a built chunk over maxBytes, the splitter halves that group and re-tests automatically.
Estimating page weights across shared fonts and images
A naive splitter that divides total file size by page count fails whenever one page holds a 15 MB scan and nine pages hold plain text. Instead, this tool walks indirect objects to amortize shared stream lengths plus 650 bytes per page and a group overhead floor of at least 1200 bytes.
Per the ISO 32000-1 PDF 1.7 Specification, every standalone PDF must carry its own catalog, cross-reference table, and copies of any fonts or logos used on its pages. If you know the exact page breaks you want instead of splitting by weight, try a manual page-range splitter.
Automatic halving when shared resources push a chunk over
After estimating initial groups, the worker serializes each chunk with PDFDocument.create() and checks its actual byte length against maxBytes (1,048,576 bytes per MB). Whenever a multi-page chunk exceeds the cap, it splits at half its length and retries until every multi-page file fits strictly under the limit.
Choosing 10 MB, 20 MB, or custom limits for email and portals
Most email providers and government portals reject attachments that exceed a fixed ceiling, even if the raw file on your disk looks slightly smaller. Picking the right Largest file size preset lets you cut a multi-page document into 10 MB pieces or smaller chunks that clear strict mail and portal caps.
| Target preset | Byte cap (1024-based) |
Safe encoded email size | Common use case |
|---|---|---|---|
| 1 MB or 2 MB | 1,048,576 to 2,097,152 bytes |
Under 2.7 MB |
Strict HR, insurance, and tax web forms |
| 5 MB | 5,242,880 bytes |
Under 6.7 MB |
Court e-filing systems and legacy portals |
| 10 MB (many email services) | 10,485,760 bytes |
Under 13.3 MB |
Corporate Exchange and Outlook servers |
| 20 MB | 20,971,520 bytes |
Under 26.6 MB |
Safer ceiling for 25 MB webmail |
| 25 MB (Gmail attachments) | 26,214,400 bytes |
Over 25 MB after Base64 |
Direct portal uploads with 25 MB raw cap |
Under the IETF RFC 2045 MIME Base64 Specification, email clients encode every 3 octets of binary PDF data into 4 ASCII characters, expanding attachments by roughly 33 percent. Choosing 10 MB or 20 MB prevents bounce errors on 25 MB mail servers, while if your document has embedded chapters, you can keep chapter boundaries intact when splitting instead.
Splitting a large PDF vs. compressing pages
Compress a PDF when you want to keep all pages in a single file by downsampling heavy images, and split by size when you must preserve full original image quality across multiple smaller files. You can also combine both workflows when a single page is too heavy on its own.
Handling single pages that already exceed the limit
If a single architectural blueprint or high-resolution scan is 14 MB on its own and your limit is 10 MB, splitting cannot divide that 1 page further. The tool places it in its own file, keeps all other chunks under 10 MB, and flags the exact page number in the result note so you can shrink oversized single pages first.
Tip: According to Google Gmail Attachment Size Limits, Gmail’s 25 MB cap applies to the entire message including headers and body text. Select 20 MB or 10 MB when emailing PDFs directly as attachments.
Downsampling scans vs. preserving lossless pages
Splitting by size never alters pixel resolution, color profiles, or vector text sharpness. Compression reduces file weight by re-encoding raster images at lower JPEG quality or lower resolution, which helps single-page uploads pass strict 2 MB portal checks.
For multi-page image-heavy packets, compressing first and then splitting by file size often cuts the total number of email parts in half while keeping every page readable on desktop and mobile screens.
Ordered ZIP output, page dimensions, and browser privacy
Multi-hundred-megabyte financial binders and medical records stay safer when split directly on your own hardware. Processing inside a local Web Worker avoids slow uploads over mobile connections and keeps private files off remote servers with 0 bytes transmitted.
Sequential filenames and zero-padded part numbers
Every output PDF is named <stem>-01-pages-1-12.pdf with zero-padded part numbers so files sort cleanly in alphabetical order inside the downloaded ZIP archive. Each filename tells you both the sequence index and the exact 1-based page span inside that chunk.
Standardizing oversized page dimensions locally
Sometimes oversized PDFs stem from poster-sized 36 x 48 inch CAD sheets mixed with standard A4 pages. After splitting out heavy sections, you can open Resize PDF to standardize mixed page dimensions before final distribution.
Keeping every part ordered inside a single ZIP download ensures recipients can reassemble or review the full document in sequence without missing pages or hitting server upload caps. Each part remains a valid standalone PDF that opens cleanly in any standard desktop, web, or mobile reader without extra software.