NewYour coding agent can read the release notes before it upgrades.Set up the MCP server →
crates.io
crates.io, #1509 most downloaded on crates.io
Quick probing of image dimensions without loading the entire file.
Latest stable
0.15.0
Last release 4 months ago
09 Jul 2026
Release timing varies
gaps range from 8 days to 1.3 years
Some releases are documented
notes for 15 of 34 stable releases
Nothing withdrawn
no release was ever pulled
10 years old
34 releases · first in 2017
Drag across the ruler to compare, or tap a release to open it.
Fix size reader for ktx2 and the test case, original code always retu… by @AllenDang in https://github.com/Roughsketch/imagesize/pull/46
Full Changelog: https://github.com/Roughsketch/imagesize/compare/v0.14.0...v0.15.0
Add features flags for each image format by @ShadowMitia in https://github.com/Roughsketch/imagesize/pull/41
Full Changelog: https://github.com/Roughsketch/imagesize/compare/v0.13.0...v0.14.0
Your coding agent can read these notes before it upgrades. Set up the MCP server →
List Avif format in README.md by @seanaye in https://github.com/Roughsketch/imagesize/pull/33
Also thanks to @virtualritz for initially introducing some formatting changes and bumping the edition.
Full Changelog: https://github.com/Roughsketch/imagesize/compare/v0.12.0...v0.13.0
Auto change log does not fully track everything since not everything went through PRs. I'll likely be more consistent with this in the future.
Auto change log does not fully track everything since not everything went through PRs. I'll likely be more consistent with this in the future.
ImageType enum so that types will now be in alphabetical orderFull Changelog: https://github.com/Roughsketch/imagesize/compare/v0.11.0...v0.12.0
Update tests so they aren't packaged with crate by @Roughsketch in https://github.com/Roughsketch/imagesize/pull/22
Full Changelog: https://github.com/Roughsketch/imagesize/compare/v0.10.1...v0.11.0
Support TIFF width and height in u16 by @gyk in https://github.com/Roughsketch/imagesize/pull/19
Full Changelog: https://github.com/Roughsketch/imagesize/compare/v0.10.0...v0.10.1
Thanks to those who submitted PRs to make this release!
Thanks to those who submitted PRs to make this release!
For this release the main changes are the following:
reader_size has been added which allows getting dimensions from anything implementing BufRead + Seek.Full Changelog: https://github.com/Roughsketch/imagesize/compare/v0.9.0...v0.10.0
This release has two changes. The main one is #13 which adds support for JXL images. This isn't a format I was aware of but @oscar-rc1 was kind enough
This release has two changes. The main one is #13 which adds support for JXL images. This isn't a format I was aware of but @oscar-rc1 was kind enough to create a PR to add it.
Behind the scenes there was also some minor refactoring which may be expanded later. The interface should stay the same but I may work on restructuring things further in the future. I doubt there will be any forward facing interface changes though.
Recently I've been pointed to a great project that also deals with image size. I was curious as to how the support for some of the formats I don't inc
Recently I've been pointed to a great project that also deals with image size. I was curious as to how the support for some of the formats I don't include works, and while looking at that I noticed there was a type of webp I had never encountered before in the wild. In order to be thorough, I've patched up this missing hole which means lossless webp images are now supported.
As for the other missing formats, out of the ones they have that this library doesn't the only reasonable candidate for inclusion would be j2c, jp2, jcx, and mng, but I have personally never seen these formats and I don't want to include something unless it's actually used.
Internally there was also some cleanup regarding the interfaces and organization of tests.
The previous implementation of JPEG parsing relied on tags being the first thing encountered in an image, and having metadata not get in the way. This
The previous implementation of JPEG parsing relied on tags being the first thing encountered in an image, and having metadata not get in the way. This was a poor implementation and led to issues where JPEG files with EXIF data could show up as corrupted or report an improper size.
As of 0.8.7 the JPEG parsing has been reworked to instead read each JPEG marker, determine whether it is the dimension tag, and if not skip the entire length of the tag. This prevents the library from getting stuck inside metadata sections and reading unrelated data as invalid markers.
After publishing 0.8.5 I was made aware that not all SOF markers are dimensions. In particular FF C4, FF C8, and FF CC do not contain valid dimension
After publishing 0.8.5 I was made aware that not all SOF markers are dimensions. In particular FF C4, FF C8, and FF CC do not contain valid dimension data. Instead of leaving this bug hidden inside, I've decided to push an update again.
I've also included a bunch of clippy fixes, but this is just to make me feel better about myself.
Previously I've only encountered JPG images using SOF0 or SOF2 (FF C0 or FF C2). However, #10 brought to my attention an SOF1 image. I looked up the S
Previously I've only encountered JPG images using SOF0 or SOF2 (FF C0 or FF C2). However, #10 brought to my attention an SOF1 image. I looked up the SOF marker definition and it looks like C0 to CF are all valid SOF markers, so I've changed the check to accept them all.
I am missing tests for C3 to CF so I can't guarantee they are all the same, but if something breaks please open an issue. As of right now all JPG test images in the repo are working.
This release fixes an issue with HEIC image parsing where the dimensions returned would be a sub-image or thumbnail and not the larger image. In order
This release fixes an issue with HEIC image parsing where the dimensions returned would be a sub-image or thumbnail and not the larger image. In order to fix this, the library will now read all spacial tags and pick the largest one as the size to return. On top of this, it now also looks for a rotation tag and will swap the width and height if the image should be rotated 90 or 270 degrees.
I've been meaning to try out fuzzing for a while, but unfortunately I only tried it after v0.8.2 was released. Running cargo fuzz found a multiplicati
I've been meaning to try out fuzzing for a while, but unfortunately I only tried it after v0.8.2 was released. Running cargo fuzz found a multiplication crash in under 10 seconds, which was very impressive and also kind of frustrating due to the recent release. This crash should not be an issue in release mode, but this does make it so debug mode does not crash anymore.
The crash involved a crafted or corrupted TIFF payload where a tag's count entry was set to a very large number. This number was eventually multiplied and converted to a usize, but since the multiplication happened as a u32, it was possible to overflow by setting the count to something like 0xFFFFFFFF. The variable is now converted to a usize after being read, so this type of overflow should not be possible anymore.
This release fixes two major issues with JPEG parsing which would cause a valid JPEG to return an error.
This release fixes two major issues with JPEG parsing which would cause a valid JPEG to return an error.
Issue one (#9) was that JPEGs that happened to have their image dimension marker be preceded by FF would be skipped (e.g., FF FF C0).
Issue two was that JPEGs where the image dimension tag was directly preceded by another tag would be skipped. (e.g., FF E1 FF C0)
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
Nothing published for this version
One column per quarter.