Initializing... drag & drop files here
Supports: TAR.XZ
.tar.xz (also written .txz) archives. Batch upload is supported — convert a folder full of source tarballs in one pass. Files process on our servers with no sign-up..tz2 per input (one-to-one conversion), or switch to Single Archive to extract every uploaded .tar.xz and re-pack the combined contents into a single .tz2.bzip2 -9 block-size 900 KB output — the same settings most Linux distros use. Leave them unless you have a specific reason to deviate; xconvert's defaults are tuned for maximum compatibility with tar xjf on any POSIX system..tz2 for download — no watermark, no email gating, no file-count cap.TAR.XZ (.tar.xz / .txz) and TZ2 (.tz2, equivalent to .tar.bz2 / .tbz2) are both tar containers with stream compression layered on top — the difference is the algorithm. TAR.XZ uses LZMA2 (XZ Utils, released 2009 by the Tukaani Project), which gives the tightest compression ratio of any mainstream Unix archiver. TZ2 uses bzip2 (Julian Seward, 1997), which is roughly 5–10% looser than xz but historically ships on every Linux, BSD, and macOS install without an extra package. Conversion is mechanical: decompress one stream, re-compress with the other, and the inner tar payload — file modes, ownership, symlinks, mtimes — passes through bit-identical.
coreutils-adjacent territory) but skip xz-utils. Repackaging as .tz2 removes the runtime dependency.bzip2recover salvage undamaged blocks from a corrupted archive, recovering most of the data. XZ's single-stream LZMA2 layout typically can't recover past the first bad byte.-9 archives..tar.bz2 for legacy reasons. If your downstream tooling expects bzip2, converting upstream is cheaper than patching the toolchain.Need a different target instead? Convert to TAR.XZ to TAR.GZ for the widest compatibility, TAR.XZ to ZIP for Windows recipients, or TAR.XZ to 7Z when you want to keep solid LZMA-class compression in a 7-Zip container. Going the other direction? See TAR.BZ2 to TAR.XZ.
| Property | TAR.XZ (.tar.xz / .txz) |
TZ2 (.tz2 / .tar.bz2 / .tbz2) |
|---|---|---|
| Outer container | POSIX tar (ustar/pax) | POSIX tar (ustar/pax) |
| Stream compressor | XZ Utils / LZMA2 | bzip2 (Burrows-Wheeler + Huffman) |
| Algorithm released | LZMA 1998; LZMA2 / .xz format 2009 |
bzip2 1997 |
| Typical ratio vs gzip | ~30–40% smaller | ~10–20% smaller |
| Compression speed | Slowest of the three (xz -9 ~10x gzip) |
Faster than xz, slower than gzip |
| Decompression speed | Faster than bzip2 at equal ratio | 100–200 MB/s typical |
| Block recovery from corruption | No (single LZMA2 stream) | Yes — bzip2recover per 900 KB block |
| Default tool flag (tar) | tar -cJf / tar -xJf |
tar -cjf / tar -xjf |
| Multithreaded by default | LZMA2 supports threads; xz -T0 opt-in |
No (use pbzip2 or lbzip2 for parallelism) |
| Memory at decompress | 8–65 MB (preset-dependent) | ~4 MB (fixed, regardless of input size) |
| First-class on minimal Linux | Needs xz-utils package |
Almost always present |
Numbers below are from a public benchmark on linux-3.18.19.tar (580 MB Linux kernel source), CentOS 7.1, 4 cores. They show the practical cost of switching from xz to bzip2 on a typical source tree.
| Metric | TAR.XZ (level 9) | TZ2 (level 9) | Difference |
|---|---|---|---|
| Output size | ~81 MB | ~97 MB | TZ2 is ~20% larger |
| Compression time | ~501 s | ~83 s | TZ2 is ~6x faster to write |
| Decompression time | ~11 s | ~24 s | TAR.XZ is ~2x faster to read |
| Memory at compress | ~700 MB | ~7 MB | xz -9 is memory-heavy |
| Memory at decompress | ~65 MB | ~4 MB | bzip2 is lighter on read |
| Recovery from byte error | None past damage | Per-block via bzip2recover |
TZ2 wins on durability |
Takeaway: TZ2 trades roughly 15–20% extra bytes for much faster compression, lighter memory use, and better corruption tolerance. TAR.XZ remains the choice when on-disk size is the only thing that matters.
.tz2 the same as .tar.bz2 and .tbz2?Yes. .tz2, .tbz2, .tb2, and .tbz are all single-extension shorthands for a tar archive compressed with bzip2 — identical bytes inside, just different filename conventions. tar -xjf file.tz2 and tar -xjf file.tar.bz2 extract the same way. Some 8.3-filesystem-era tooling and certain Windows decompressors prefer the three-letter .tz2 form; modern Linux distros use .tar.bz2. You can rename freely; the content doesn't change.
No. Both formats are lossless and both wrap the same POSIX tar payload. The conversion decompresses the LZMA2 stream and re-compresses with bzip2 — file contents, permissions, ownership, symlinks, hardlinks, mtimes, and extended attributes (if encoded in pax headers) pass through bit-identical. The only thing that changes is the outer compressed stream's bytes and the resulting file size.
For most real-world data — source code, log files, mixed binaries — expect the .tz2 to be 15–25% larger than the .tar.xz. Highly redundant data (text logs, repeated boilerplate) closes the gap to under 10%. Already-compressed payloads (JPEG, MP4, pre-zipped libraries) bunched into the tar make both archives nearly the same size, since neither algorithm can squeeze entropy that's already maxed out.
bzip2 isn't obsolete; it's still actively maintained (bzip2 1.0.8 released 2019, with bzip2-1.1 work ongoing) and remains the default j flag in GNU tar. Reasons to pick TZ2 today: (1) better resilience against xz-utils supply-chain incidents like CVE-2024-3094, (2) lower decompression memory on embedded systems, (3) block-level recoverability from corruption, and (4) zero install footprint on minimal Linux/BSD images that ship libbz2 but not xz-utils. If none of those apply, .tar.xz is the better choice on size alone.
macOS: yes — the built-in tar (BSD tar) handles bzip2 transparently via tar -xf file.tz2. Windows 11: yes since the May 2023 update, Windows Explorer extracts .tar.bz2 natively, and PowerShell's tar (shipped since Windows 10 1803) accepts .tz2 directly. Older Windows needs 7-Zip, WinRAR, or PeaZip — all free, all read bzip2 tarballs. Linux distros ship bzip2 and tar in every base install.
No — this tool re-wraps the existing tar payload in a different compression stream; it doesn't open or rewrite the tar contents themselves. If you need to add, remove, or inspect individual files, extract the .tar.xz first, modify with tar --append / tar --delete, then convert. xconvert preserves the original tar member list, ordering, modes, and ownership exactly as the input archive recorded them.
TGZ (.tar.gz) uses gzip — fastest compression and decompression of the three, but worst ratio (typically 20–35% larger than .tz2 on text data). TZ2 sits in the middle: ~15% smaller than .tar.gz for source code, with corruption recovery support gzip lacks. Pick .tz2 for archival where size matters but you still want bzip2's recoverability; pick .tar.gz for build artefacts you'll re-fetch frequently or pipe over slow CPUs. See TAR.XZ to TGZ if you want gzip output instead.
.tz2 work with tar -xjf on every Linux distribution?Yes. The j flag (or auto-detection via tar -xf file.tz2 on GNU tar 1.15+) invokes bzip2 decompression. Every mainstream Linux (Debian, Ubuntu, Fedora, Arch, Alpine, openSUSE, Slackware), every BSD (FreeBSD, OpenBSD, NetBSD), Solaris/illumos, and macOS includes bzip2 and libbz2 in the base system or core packages. The only common environments without bzip2 are extremely stripped containers — and in those cases you can statically link bzip2 at under 80 KB.
.tz2 smaller than .tb2?They're identical. .tz2, .tb2, .tbz, .tbz2, and .tar.bz2 are five spellings of the exact same file format. Renaming a file from .tz2 to .tb2 changes nothing inside. See TAR.XZ to TB2 if you want the alternate extension on the output.