Idle Auto-Stop
Walk away from a GPU and it warns you, saves your environment, and stops the box — so an instance you forgot about stops billing instead of running all weekend.
Provider coverage
The toggle is available on every provider — its only requirement is a selected storage bucket, not a specific provider. Idle detection reads the same utilization data as Live GPU Metrics, so a GPU profile that can't report utilization (for example, Vultr's vGPU-based A16 line) shows as "we can't read utilization on this box" instead of being guessed at.
On by default when you deploy from the console
If you deploy from the console with a storage bucket selected, Auto-save & stop when idle is already switched on in the deploy sheet. Leave it on and the deployment is watched from the moment it comes up; switch it off there, or later from that deployment's Idle tab, and nothing is watched or stopped.
It is per deployment, never account-wide, and it is never switched on retroactively for boxes you already have running. A deploy made through the API or the aq CLI has it off unless you set the policy explicitly. It also needs a storage bucket: it saves your environment before stopping it, and without somewhere to write that save it can't run, so the toggle stays disabled until you pick one.
Left at their defaults, the thresholds are 5% GPU utilization, a warning at 30 minutes idle, and save-then-stop at 60 minutes.
Warn, then save and stop
Set a GPU utilization threshold that counts as idle, how many minutes of that before you're warned, and how many minutes before it acts. Utilization at or under the threshold for the warn period sends an email — the box is still running and still billing at that point. Staying idle past the act threshold saves the box first, then stops it.
The warn threshold always has to sit below the act threshold; the console won't save a configuration where the warning would fire after the box already stopped.
Four states, and unknown is a real one
A deployment with auto-stop configured shows one of: in use, idle, idle past the auto-stop threshold, or we can't read utilization on this box. That last one is not a rounding of "probably idle" — it's an explicit unknown, shown whenever utilization or on-box presence can't be read, and it never triggers a stop.
Presence is read from the on-box agent, which sees SSH logins and browser-terminal sessions. The system doesn't infer that nobody is on a box from silence — an unreadable signal stays unknown rather than being treated as idle.
FAQ
Is idle auto-stop on by default?
In the console, yes — the deploy sheet's "Auto-save & stop when idle" toggle is pre-checked whenever a storage bucket is selected, so a console deploy is watched unless you switch it off. It is per deployment rather than account-wide, and it is off by default for deploys made through the API or the aq CLI unless you enable it explicitly. You can turn it off at deploy time or afterward from the deployment's Idle tab.
What happens right before it stops the box?
It saves your environment to the configured storage first, the same save mechanism used for a manual stop-with-save, then stops the box. It's a file save, not a resume point for whatever was running in memory.
What does "we can't read utilization" mean?
It means the platform genuinely can't tell whether the box is idle or in use — the utilization reading or the on-box presence signal is unavailable. It's shown as its own state and never treated as idle, so it can't trigger an auto-stop on a guess.
Can I change the thresholds after deploying?
Yes. Open the deployment's Idle tab to turn auto-stop on or off, or change the utilization threshold, warn minutes, and act minutes at any time.
Does it work the same on every GPU?
The toggle itself is available regardless of provider. Idle detection reads the same utilization data as Live GPU Metrics, so a GPU whose virtualization profile can't report utilization shows as unknown rather than being marked idle.
