Choose a version:
33% The original file has 40529 bytes (39.6k) and is available from the project website.
There you can find the official minified version, too, which brings down the size to 13403 bytes (13.1k, 33%).

After GZIP compression these minified files vary in size:
Boot
  5140 bytes (5.0k)
CDN
gzip -6 (default)
  4643 bytes (4.5k)
local copy
Baidu
  4637 bytes (4.5k)
CDN
cdnjs
  4635 bytes (4.5k)
CDN
gzip -9
  4626 bytes (4.5k)
local copy
unpkg
  4615 bytes (4.5k)
CDN
zultra
  4509 bytes (4.4k)
local copy
7zip -mx=9 -tgzip
  4506 bytes (4.4k)
local copy
libdeflate -12
  4502 bytes (4.4k)
local copy
kzip -s0 -rn -b0
  4499 bytes (4.4k)
local copy
pigz -11 -n
  4494 bytes (4.4k)
local copy
Zopfli
  4488 bytes (4.4k)
local copy

perma-link to the smallest file on my server:
http://minime.stephan-brumme.com/files/underscorejs/underscore-1.4.0.min.js (or via HTTPS)

You will automatically get the smallest UnderscoreJS 1.4.0 file, ETag caching is available and
if your browser doesn't support GZIP decompression then the uncompressed version will be sent.

Currently best Zopfli settings

Save 127 bytes by using my UnderscoreJS 1.4.0 Zopfli version instead of the best available CDN (2.83% smaller than unpkg, 4488 vs. 4615 bytes):
You can use my super-compressed files for whatever purpose you like as long as you respect the library's original license agreement.
There are no restrictions from my side - but please avoid hot-linking if you run a high-traffic website.

These command-line settings yielded the best compression ratio so far (Linux version of zopfli-krzymod):
zopfli --i100000 --mb8 --mls512 --bsr10 --lazy --ohh

(found February 12, 2019)
Description Value Parameter
iterations 100000  --i100000
maximum blocks 8  --mb8
maximum length score 512  --mls512
block splitting recursion 10  --bsr10
lazy matching in LZ77 yes  --lazy
optimized Huffman headers yes  --ohh
initial random W for iterations 1  --rw1
initial random Z for iterations 2  --rz2

Verify file integrity

After decompression, my uncompressed files are identical to the original ones:

MD5:
curl --silent --compressed https://raw.githubusercontent.com/jashkenas/underscore/1.4.0/underscore-min.js --location | md5sum
51b45e9de2a9359e0d17e94e671e829a  -
curl --silent --compressed https://minime.stephan-brumme.com/files/underscorejs/underscore-1.4.0.min.zopfli.js.gz | md5sum
51b45e9de2a9359e0d17e94e671e829a  -

SHA1:
curl --silent --compressed https://raw.githubusercontent.com/jashkenas/underscore/1.4.0/underscore-min.js --location | sha1sum
7fdcf03767901f9aad28dcd11e97bcb96f51c4b9  -
curl --silent --compressed https://minime.stephan-brumme.com/files/underscorejs/underscore-1.4.0.min.zopfli.js.gz | sha1sum
7fdcf03767901f9aad28dcd11e97bcb96f51c4b9  -

These CDNs send you the original file:
CDN Size (compressed) MD5 (uncompressed) Timestamp
Boot 5140 bytes 51b45e9de2a9359e0d17e94e671e829a (invalid)
cdnjs 4635 bytes 51b45e9de2a9359e0d17e94e671e829a (invalid)
unpkg 4615 bytes 51b45e9de2a9359e0d17e94e671e829a July 11, 2016 @ 16:49

And some CDNs send you a different file:
CDN Size (compressed) MD5 (uncompressed) Comment / Diff Timestamp
Baidu 4637 bytes 50117e8f4bc2072dff5dab939bd3269e only whitespaces differ (invalid)

Note: only the MD5 hashes are shown to keep things simple.

Other Versions

Available UnderscoreJS versions at minime.stephan-brumme.com:

1.13.6, 1.13.5, 1.13.4, 1.13.3, 1.13.2, 1.13.1, 1.13.0,
1.12.1, 1.12.0,
1.11.0,
1.10.2, 1.10.1, 1.10.0,
1.9.2, 1.9.1, 1.9.0,
1.8.3, 1.8.2, 1.8.1,
1.7.0,
1.6.0,
1.5.2, 1.5.0,
1.4.4, 1.4.3, 1.4.2, 1.4.1, 1.4.0,
1.3.3, 1.3.2, 1.3.1, 1.3.0,
1.2.4, 1.2.3, 1.2.2, 1.2.1, 1.2.0,
1.1.7, 1.1.5, 1.1.4, 1.1.3, 1.1.2, 1.1.1, 1.1.0,
1.0.4, 1.0.3, 1.0.2, 1.0.1, 1.0.0,
0.6.0,
0.5.7, 0.5.6, 0.5.5, 0.5.4, 0.5.3, 0.5.2, 0.5.1, 0.5.0,
0.4.7, 0.4.6, 0.4.5, 0.4.4, 0.4.3, 0.4.2, 0.4.1, 0.4.0,
0.3.3, 0.3.2, 0.3.1, 0.3.0,
0.2.0,
0.1.1, 0.1.0

The project site contains an overview how well these versions were compressed.
Other interesting projects are AngularJS, BackboneJS, Bootstrap, D3, Dojo, Ember, jQuery, Knockout, lodash, React, Socket.IO, ThreeJS and Vue.

Changelog

Best Zopfli parameters so far:
Size Improvement Parameters Found
4488 bytes -1 byte zopfli --i100000 --mls512 --bsr10 --lazy --ohh February 12, 2019 @ 22:39
4489 bytes -1 byte zopfli --i10000 --mls512 --bsr9 --lazy --ohh September 8, 2015 @ 14:47
4490 bytes -1 byte zopfli --i1000 --mls512 --bsr9 --lazy --ohh September 8, 2015 @ 10:52
4491 bytes zopfli --i100 --mls8192 --bsr4 --lazy --ohh September 8, 2015 @ 10:33

If there are multiple parameter sets yielding the same compressed size, only the first one found is shown.

Most recent activity on July 20, 2020 @ 12:54.

Heatmaps

This Zopfli heatmap visualizes how compression changes when modifying the --bsr and --mls parameter.
Cell's contents is the best filesize achieved (in bytes, hover with mouse over cells to see number of iterations).

Good parameters are green, bad are red. The best and worst are bold as well.
The brightness of the blue background color indicates how many iterations were processed:
10,000, 100,000 or 1,000,000.
bsr \ mls
2 4 8 16 32 64 128 256 512 1024 2048 4096 8192 16384 32768
bsr \ mls
2 4 8 16 32 64 128 256 512 1024 2048 4096 8192 16384 32768
4497 4497 4491 4490 4491 4490 4498 4491 4490 4491 4497 4493 4491 4492 4492
4495 4495 4496 4495 4494 4495 4493 4491 4494 4492 4493 4490 4488 4491 4491
4494 4494 4495 4493 4494 4491 4491 4491 4493 4491 4492 4490 4489 4491 4491
4490 4490 4494 4493 4493 4493 4493 4490 4491 4491 4493 4491 4497 4491 4491
4491 4491 4491 4489 4491 4495 4495 4490 4489 4492 4491 4488 4491 4492 4492
4495 4495 4495 4495 4493 4491 4494 4488 4489 4489 4491 4490 4490 4491 4491
4495 4495 4490 4489 4491 4490 4493 4490 4488 4491 4493 4490 4490 4490 4490
4490 4490 4499 4491 4491 4493 4493 4490 4490 4489 4493 4490 4488 4490 4490
4497 4497 4490 4491 4491 4491 4491 4490 4491 4491 4492 4489 4491 4491 4491
4497 4497 4489 4491 4490 4491 4493 4491 4491 4492 4491 4489 4489 4489 4489
4499 4499 4490 4489 4494 4493 4493 4488 4491 4492 4491 4490 4488 4491 4491
4493 4493 4491 4494 4491 4492 4493 4488 4489 4489 4492 4490 4492 4491 4491
4493 4493 4493 4489 4491 4490 4494 4496 4491 4491 4488 4496 4491 4491 4491
4491 4491 4495 4491 4491 4495 4491 4490 4491 4491 4490 4491 4491 4490 4490
4491 4491 4490 4491 4491 4491 4491 4488 4489 4490 4488 4490 4492 4492 4492
4495 4495 4493 4493 4494 4489 4493 4490 4491 4491 4493 4489 4492 4492 4492
4493 4493 4493 4494 4494 4490 4493 4490 4489 4491 4493 4489 4492 4491 4491
4490 4490 4491 4494 4494 4491 4493 4490 4491 4491 4491 4490 4490 4491 4491
4490 4490 4491 4489 4494 4490 4491 4490 4491 4491 4493 4490 4492 4491 4491
4490 4490 4491 4491 4491 4491 4493 4492 4491 4491 4488 4490 4492 4490 4490
4494 4494 4494 4492 4490 4490 4492 4488 4491 4491 4488 4489 4492 4490 4490
4493 4493 4490 4491 4494 4490 4493 4490 4492 4492 4491 4489 4491 4491 4491
4495 4495 4494 4494 4490 4489 4492 4488 4492 4491 4492 4489 4492 4491 4491

Due to the Monte Carlo design of my search algorithm, not all parameters have reached the same number of iterations yet:
Iterations Min. Bytes Reduction Coverage
100 4491 bytes 100%
1,000 4490 bytes -1 byte 100%
10,000 4489 bytes -1 byte 100%
100,000 4488 bytes -1 byte 12.17%
1,000,000 4488 bytes 0.29%
10,000,000

KZIP has far less options available for tuning/optimization. I only played around with the number of blocks (parameter -n):
Blocks Min. Bytes Compared To Best Zopfli Compared To Best KZIP
4499 bytes +11 bytes (+0.25%)
4499 bytes +11 bytes (+0.25%)
4518 bytes +30 bytes (+0.67%) +19 bytes
4518 bytes +30 bytes (+0.67%) +19 bytes
4550 bytes +62 bytes (+1.38%) +51 bytes
4585 bytes +97 bytes (+2.16%) +86 bytes
4621 bytes +133 bytes (+2.96%) +122 bytes
4650 bytes +162 bytes (+3.61%) +151 bytes
4681 bytes +193 bytes (+4.30%) +182 bytes

Non-DEFLATE Algorithms

Archivers based on completely different compression algorithms often produce superior results.
Unfortunately, browsers only support gzip compression at the moment.
However, support for Brotli is constantly growing - but your browser doesn't support it.
Algorithm Program Parameters Size Compared To Best Zopfli
RAR (proprietary) RAR rar a -m5 -md64m -mc63:128t -mt1 4005 bytes -483 bytes (-10.76%)
Brotli (Wikipedia) brotli brotli -q 11 4173 bytes -315 bytes (-7.02%)
PPMd (Wikipedia) 7zip 7za a -mx=9 -m0=ppmd 4405 bytes -83 bytes (-1.85%)
Zstandard (Wikipedia) zstd zstd -19 4512 bytes +24 bytes (+0.53%)
Burrows-Wheeler transform (Wikipedia) bzip2 bzip2 -9 4526 bytes +38 bytes (+0.85%)
LZMA2 (Wikipedia) xz xz -9 4580 bytes +92 bytes (+2.05%)
ZPAQ (Wikipedia) zpaq zpaq -method 69 4857 bytes +369 bytes (+8.22%)

Detailled Analysis

I wrote a DEFLATE decoder in Javascript. Click the button below to start a client-side analysis of the smallest gzipped files (may take a second):


Notes: pigz is a fast open source multi-threaded implementation of gzip written by one of the original authors of gzip.
However, when using compression level 11, pigz actually switches to the slower Zopfli algorithm and isn't multi-threaded anymore.
KrzyMOD's extensions to Zopfli offer the highest level of configuration and is therefore used for my brute-force search.
Ken Silverman wrote the closed-source KZIP compression program and Jonathon Fowler ported it to Linux.
Defluff was created by Joachim Henke; DeflOpt is a tool by Ben Jos Walbeehm.

website made by Stephan Brumme in 2015 and still improving in 2024.
all timestamps are displayed in central european time. see my changelog.
no flash, not even images or external css files - and everything squeezed into a single html file.
which was handsomely compressed before releasing it into the wild internet - obviously.

please visit my homepage and my blog, too.
email: minime (at) stephan-brumme.com

All trademarks are property of their respective owners. You know, the boring legal stuff.