Runners
Hosted runners
0 runners online in the shared poolYour runs are executed by our runners unless one of your own is online. This month: 0 of 40 hosted site-runs used (resets on the 1st, UTC). Hosted runners only test verified sites (0 of 6 covered by a verified domain). A run is split: verified sites go to hosted runners, private hosts and unverified sites wait in their own run for your runner.
Connect your own runner (optional)
Your own runner takes precedence over the pool, has no monthly limit, and keeps test accounts and screenshots on your machine.
A runner is the open-source QA.pro engine on a machine you control. It needs docker, python 3 and git, plus an ollama endpoint with a vision model (on the same machine or another one). It pulls runs from here and uploads the results.
One line, Linux or macOS
QAPRO_ENROLL=<ask-an-admin-for-the-token> sh -c "$(curl -fsSL https://qa.pro/install.sh)" -- --server https://qa.pro --name "$(hostname)"One line, Windows PowerShell
irm https://qa.pro/install.ps1 | iex; Install-QAproRunner -Server https://qa.pro -Enroll <ask-an-admin-for-the-token>Add --ollama http://gpu-box:11434 (-Ollama on Windows) when the vision model runs on another machine. Both scripts need docker, python 3 and git;
they fetch the engine, build the browser image and start the runner. Already have a checkout? QAPRO_ENROLL=<ask-an-admin-for-the-token> python3 ee/runner/qapro_runner.py --server https://qa.pro
Only team admins see the enrollment token; ask one for it.
2 runners
| Name | Version | Last seen | State | actions |
|---|---|---|---|---|
| office-gpu | 0.7.5 | 27 Sep 2026, 01:59 UTC | online | |
| hetzner-gpu-1 | 0.7.5 | 27 Sep 2026, 01:59 UTC | online |