# Performance difference when running from sources

**URL:** <https://forum.openrefine.org/t/performance-difference-when-running-from-sources/2710>\
**Category:** Support and Helpdesk\
**Created:** [January 18, 2026, 5:18pm UTC](https://forum.openrefine.org/t/performance-difference-when-running-from-sources/2710 "2026-01-18T17:18:19Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![Jan](https://dub1.discourse-cdn.com/flex017/user_avatar/forum.openrefine.org/jan/32/48_2.png) [@Jan](https://forum.openrefine.org/u/Jan)\
**Post date:** [January 18, 2026, 5:18pm UTC](https://forum.openrefine.org/t/performance-difference-when-running-from-sources/2710/1 "2026-01-18T17:18:19Z")

</div>

hey,

when importing a larger set of data (csv) I noticed a huge difference when I build and run OpenRefine from sources vs when I use the official release build. With some files, the difference is seconds vs minutes

I checked out the tag `3.9.5` and compare it to the `3.9.5` .dmg that I downloaded.

My memory settings are the same on both version so I wonder why there is such a big difference.

Could that be because of the JRE which is `11.0.28+6` in the bundled OpenRefine? Locally, I use `21.0.7`

Below is a recording of both imports .. I know it’s not a very precise measurement but it’s just for illustration purposes. the difference is actually quite noticable

**Bundled OpenRefine**

![bundled](https://europe1.discourse-cdn.com/flex017/uploads/openrefine/original/2X/b/b1d54a34efd7b07381e5f785146ee8af3cb05f68.gif)

**Open Refine built from sources**

![local](https://europe1.discourse-cdn.com/flex017/uploads/openrefine/original/2X/9/923bcd2d246b4202a80f27d7b215a9da4c70e758.gif)