How to multi-part archive creator online (browser-native)
- Step 1Open the tool — Go to /archive-tools/multi-part-archive-creator. It's a Pro-tier tool, so sign in on a Pro plan or higher; you'll see a Pro overlay otherwise.
- Step 2Drop your files — Drag one or more local files onto the dropzone (it accepts multiple). This tool builds a fresh ZIP from your input — it does not read an existing archive. To split an archive you already have, use the Archive Splitter instead.
- Step 3Set the part size — Type a Part size in MB. Default is 25 (Gmail-safe). The field clamps to 1–4096; one MB here means 1,000,000 bytes (decimal), so a 25 MB part is exactly 25,000,000 bytes.
- Step 4Generate — Click Generate. fflate compresses every file into one ZIP at level 6, then the tool slices that ZIP into ceil(totalBytes / partBytes) ordered parts.
- Step 5Download every part — The result panel lists each part (archive.zip.001, .002, …) with its size and a 'Download N parts' button. Save all of them into the same folder — you need every part to reassemble.
- Step 6Reassemble on the other side — Concatenate in order then unzip: on Unix run cat archive.zip.* > archive.zip && unzip archive.zip; on Windows run copy /b archive.zip.001+archive.zip.002 archive.zip. Many unzippers also open archive.zip.001 directly.
Controls, defaults, and real behaviour
The only exposed control is Part size (MB). Everything else — compression level, part naming, output format — is fixed by the implementation and described accurately below.
| Control / behaviour | Value | Notes |
|---|---|---|
| Part size (MB) | Number input, default 25, min 1, max 4096 | 1 MB = 1,000,000 bytes (decimal). A 25 MB part = 25,000,000 bytes, which clears Gmail's 25 MB cap |
| Compression level | Fixed at DEFLATE level 6 | There is no level slider on this tool. To compare levels use compression-level-optimizer |
| Input | One or more local files (generative) | Builds a new ZIP from your files. It does not open an existing ZIP/7z/tar |
| Part naming | archive.zip.001, archive.zip.002, … | 3-digit, zero-padded, 1-indexed; sorts and concatenates in order |
| Output type | Split set of byte-parts (zip-split) | A raw byte split of one ZIP — not a ZIP64 multi-volume archive |
| Number of parts | ceil(totalZipBytes / partBytes) | If the ZIP is smaller than one part, you get a single archive.zip.001 |
Part size presets for common targets
Type any of these into the Part size field. There is no dropdown of presets — the field is a free number from 1 to 4096.
| Target | Suggested part size (MB) | Why |
|---|---|---|
| Gmail attachment | 25 | Gmail rejects attachments over 25 MB; this is the tool's default |
| Outlook.com attachment | 33 | Outlook.com's web limit is ~33 MB per message |
| Typical enterprise mail | 25 | Most corporate mail gateways cap at 25 MB — stay safe at the default |
| FAT32 USB stick | 4000 | FAT32 can't hold a single file ≥ 4 GiB; 4000 MB parts stay comfortably under that |
| Discord (free upload) | 8 | Discord's free per-file limit is 8 MB — set 8 to fit each part |
| Chunked/resumable upload form | 5–50 | Smaller parts retry faster on flaky connections; pick to taste |
Tier limits (archive family)
This is a Pro-minimum tool, so the Free row is shown only for reference. Limits are per archive: a byte size cap, a per-archive entry-count cap, and a batch file count.
| Tier | Max input size | Entry limit | Files per job |
|---|---|---|---|
| Free | 50 MB | 500 | 1 (tool requires Pro) |
| Pro | 500 MB | 50,000 | 20 |
| Pro + Media | 2 GB | 500,000 | 100 |
| Developer | 2 GB | 500,000 | unlimited |
Cookbook
Real before/after sketches showing what the tool produces and how recipients rejoin the parts. Filenames are illustrative.
Split a 60 MB report bundle for Gmail
Three files total ~60 MB after zipping. With the default 25 MB part size, the tool emits three parts that each fit a Gmail attachment.
Input files (dropped): q4-report.pdf 41 MB appendix.xlsx 14 MB cover-letter.docx 0.3 MB Part size: 25 Output parts: archive.zip.001 25.0 MB archive.zip.002 25.0 MB archive.zip.003 ~5.7 MB Attach one part per email — all three clear Gmail's 25 MB cap.
Reassemble on Unix
The recipient drops all parts in one folder and concatenates them in order. The glob expands in sorted order because the parts are zero-padded.
$ ls archive.zip.001 archive.zip.002 archive.zip.003 $ cat archive.zip.* > archive.zip $ unzip archive.zip Archive: archive.zip inflating: q4-report.pdf inflating: appendix.xlsx inflating: cover-letter.docx
Reassemble on Windows
Windows has no cat; copy /b concatenates in binary mode. List the parts explicitly (or use the wildcard) to control order.
C:\> copy /b archive.zip.001+archive.zip.002+archive.zip.003 archive.zip
3 file(s) copied.
C:\> tar -xf archive.zip (Windows 10+ ships bsdtar)
Many GUI tools (7-Zip, WinRAR, Keka) also open archive.zip.001 directly.Smaller file than one part
If everything zips down below your part size, you get a single part. That's expected — it's still named .001 so the reassembly recipe is identical.
Input: notes.txt (12 KB) Part size: 25 Output: archive.zip.001 ~8 KB Reassembly is the same one-liner; with one part it's just a rename/extract.
8 MB parts for a Discord upload
Discord's free tier caps each upload at 8 MB. Set Part size to 8 and post each part as a separate attachment; the recipient rejoins with cat.
Part size: 8 Output parts: archive.zip.001 8.0 MB archive.zip.002 8.0 MB archive.zip.003 ~3.4 MB Upload .001/.002/.003 to the channel, then: cat archive.zip.* > archive.zip
Edge cases and what actually happens
You dropped an existing ZIP to split it
Wrong toolThis tool builds a ZIP from local files — it does not split an archive you already have. If you drop a finished .zip it will be packed inside a new ZIP and then split, which is rarely what you want. To split an existing archive by size or entry count, use archive-splitter.
A recipient downloaded only some parts
Reassembly failsEvery part is required. Because the output is one ZIP cut at arbitrary byte boundaries, a missing middle part leaves a gap and the concatenated file won't unzip. Re-send the missing part — there's no per-part recovery.
Parts concatenated out of order
Reassembly failsOrder matters: archive.zip.001 must come first. The zero-padded names sort correctly with cat archive.zip.* and copy /b archive.zip.*, but if you list them by hand make sure .001 leads. Out-of-order bytes produce an invalid ZIP.
Free tier user opens the tool
Pro requiredMulti-Part Archive Creator has minTier Pro. On the Free plan you'll see a Pro overlay instead of the dropzone. Upgrade to Pro to use it (and to get the 500 MB / 50,000-entry caps).
Total bundle exceeds your tier size cap
Tier limit exceededPro caps input at 500 MB, Pro + Media and Developer at 2 GB. The combined size of your dropped files is what counts, not the per-part size. Trim the bundle or upgrade the tier.
You set Part size to 0 or left it blank
ClampedThe processor applies Math.max(1, splitSizeMb), so anything below 1 becomes 1 MB. You can't produce zero-byte or negative parts.
Already-compressed inputs (JPEG, MP4, ZIP)
ExpectedDEFLATE level 6 barely shrinks entropy-dense files, so the resulting ZIP — and therefore the part count — stays close to the raw total. This is normal; the value here is splitting, not extra compression.
You expected a true multi-volume ZIP (.z01/.z02)
By designThe tool produces a raw byte split (archive.zip.001 …), not the spanned .z01/.z02 format some tools emit. This is deliberate — byte-split parts rejoin with plain cat/copy on any OS, with no ZIP64 spanning-support surprises.
Part name collides with an existing download
Browser-handledIf archive.zip.001 already exists in your Downloads folder, the browser appends ' (1)' to the new file. Rename back to the .001/.002 sequence before reassembling, or download into a clean folder.
You need each part encrypted
Use a different tool firstThis tool does not encrypt. To protect the contents, build an AES-256 ZIP with encrypted-zip-creator first, then split that — though note this tool re-zips files, so for true 'encrypt then split' you'd split the encrypted ZIP with archive-splitter.
Frequently asked questions
Does this tool upload my files?
No. The files are read locally with the File API and compressed by fflate WebAssembly inside your browser tab. Open DevTools → Network during a run and you'll see no outbound request carrying your data.
What part size should I use for email?
Start with the default 25 MB — it clears Gmail's 25 MB attachment cap. Outlook.com allows ~33 MB. Most corporate mail gateways cap at 25 MB, so the default is the safe choice.
How do recipients rejoin the parts?
Put every part in one folder, then concatenate in order: cat archive.zip.* > archive.zip on Unix/macOS, or copy /b archive.zip.001+archive.zip.002 archive.zip on Windows. Then unzip the result. Many unzippers also open archive.zip.001 directly.
What are the parts named?
archive.zip.001, archive.zip.002, … — three-digit, zero-padded, starting at 001. The padding makes them sort and glob in the correct order.
Is one MB 1,000,000 or 1,048,576 bytes here?
1,000,000 bytes (decimal). The processor computes partBytes = partMb × 1,000,000, so a 25 MB part is exactly 25,000,000 bytes — comfortably under Gmail's 25 MiB cap.
Can I choose the compression level?
Not on this tool — it's fixed at DEFLATE level 6, the speed/size sweet spot. To compare levels 1/3/6/9 on your data, use the Compression Level Optimizer.
Can I split a ZIP I already have?
Not with this tool — it builds a new ZIP from local files. To split an existing archive (by size or by entry count) use the Archive Splitter instead.
Why isn't this a true multi-volume ZIP?
True multi-volume (.z01/.z02) relies on ZIP spanning support that isn't universal. A raw byte split rejoins with a one-line cat/copy on any platform, so it's the more portable choice.
What's the maximum bundle size?
Pro: 500 MB. Pro + Media: 2 GB. Developer: 2 GB. That's the total size of your dropped files; the per-part size is independent.
Is this free?
The tool itself requires a Pro plan or higher (minTier Pro). On the Free plan you'll see a Pro overlay.
Will the parts open in 7-Zip / WinRAR / Keka?
Yes — point them at archive.zip.001 and most will auto-detect the sequence, or concatenate first. The reassembled file is a standard ZIP.
What if my bundle is smaller than one part?
You'll get a single archive.zip.001. That's expected; the reassembly recipe still works (it's effectively a rename/extract).
Privacy first
Every JAD Archive tool runs entirely in your browser using fflate, @zip.js/zip.js, and the libarchive WASM bridge. Your archives never leave your device — verified by zero outbound network requests during processing.