Chatwoot, Zammad, FreeScout, UVdesk: open-source Zendesk alternatives by repo data

My open-source alternatives directory's Zendesk page lists four repositories. The numbers below are from the 2026-10-18 refresh. Stars, forks, language and license come from saas.json. Release data and open items come from evidence.json, generated at 06:03 UTC that day.

The table

"Open items" is GitHub's open issue count, which also counts open pull requests.

None of the four is archived. All four were pushed between October 7 and October 18.

Chatwoot's license isn't one word

GitHub's API reports Chatwoot's license as NOASSERTION, which means it couldn't match the file to a single SPDX ID. My directory has a manual override for it, checked against the LICENSE file on 2026-09-05. Everything outside enterprise/ is MIT. The enterprise/ directory has its own separate license. If you plan to self-host and modify it, read enterprise/LICENSE before you assume "MIT". I've written about why the directory can't print some GitHub license values verbatim. This repo is one of those cases.

The other three are more straightforward. Zammad and FreeScout are AGPL-3.0, so if you modify them and run them as a network service, you have to share your changes. UVdesk uses OSL-3.0, which you don't see often. It's also a copyleft license, so check it with whoever handles licensing for you.

Releases: three different stories

FreeScout published 54 stable GitHub releases in the last 12 months. The latest, 1.8.248, came out the day before the refresh, and it has the fewest open items of the four at 20.

Chatwoot published 23 releases, the latest on September 18. Its 1,600 open items is the highest count here. It's also the most-forked repo by far.

Zammad has no stable GitHub releases in my data (the ETL skips drafts and prereleases), so its zero isn't a measure of how often it ships. It was pushed on October 16. Release counts measure how a project tags, not how often it ships, and this is a clean example. A "releases per year" column would rank Zammad last for a reason that has nothing to do with maintenance.

UVdesk did publish releases, and then stopped. The last release was v1.1.8 in September 2025, so none in the past 12 months, but the repo was pushed on October 7. Something is changing there. It just isn't getting tagged.

One ratio that stands out

Most repos in this group have 4 to 7 stars per fork: Chatwoot about 4.1, Zammad about 5.6, FreeScout about 6.3. UVdesk's community-skeleton has about 34.9 (19,636 stars, 563 forks). I'm not going to guess why. A skeleton repo may simply get forked less than the main codebase of a product. But it means stars alone overstate how UVdesk compares to the others on any measure that involves people building on the code.

GitHub also labels UVdesk's language as CSS. The directory's notes describe it as PHP-based. GitHub's language detection only counts bytes, so treat that column as a rough hint, not a spec.

How I'd read it

On these numbers, FreeScout is shipping most actively with the smallest backlog, and Chatwoot has by far the biggest following. Zammad's GitHub data can't tell you much about how often it ships, so you'd need to look at its own release channel. UVdesk's release history has stalled even though the repo is still being pushed to.

None of this measures support features. The directory's editorial notes say Zammad is closest to Zendesk's ticketing with SLA policies, and Chatwoot is strongest on live chat and omnichannel messaging. Those are judgments, not measurements, so run them past a trial install. For the same kind of comparison on internal-tool builders, see ToolJet, Appsmith, Budibase and Lowdefy.

I left out a "What real users say" section on purpose. I couldn't collect a comment sample for these four this week, and I won't write one without counted, linkable sources.

Part of an ongoing 6-month experiment running three AI-curated directory sites. The technical claims here are real; this article was AI-assisted.