Catch idle CoreWeave GPUs before they waste budget
Every hour, WebRun opens the CoreWeave cloud dashboard, reads utilization and running time for every GPU instance, saves a dated snapshot to Google Drive, and posts a Slack alert naming any instance that's been idle for more than 2 hours.
- No credit card
- Under $0.01 per run
- Cancel anytime
How do I catch idle CoreWeave GPU instances before they waste budget?
WebRun checks every CoreWeave GPU instance on a regular schedule, reading utilization and how long each one has been running. It saves a usage snapshot to Google Drive and posts a Slack alert naming any instance that's been idle for more than 2 hours, so a forgotten or crashed job stops quietly burning through GPU budget.
- Idle GPUs get flagged within hours instead of running up a bill unnoticed
- A running utilization history builds up in Drive automatically
- The team gets an alert only when a GPU is genuinely idle
Built for ML engineering teams · AI startups · platform engineers · research teams
What does WebRun do on every run?
The exact actions WebRun takes, in order - in plain language, so you can adjust anything.
-
WebRun signs in and gets to work
Opens
cloud.coreweave.comin a real browser with your saved login - no setup, no API keys. -
1
CoreWeave - check GPU utilization
coreweave.com
WebRun opens CoreWeave to check GPU utilization. - Open the CoreWeave cloud dashboard and review every GPU instance
- Read current utilization and how long it's been running
- Flag any instance under meaningful utilization for more than 2 hours
Done when Every GPU instance has a current utilization reading.
-
2
Google Drive - save the usage snapshot
drive.google.com How to Automate Google Drive
WebRun opens Google Drive to save the usage snapshot. - Save the utilization snapshot as a dated file in your shared Drive folder
- Keep prior snapshots for a running usage history
Done when This run's snapshot is saved to Drive.
-
3
Slack - alert on idle GPUs
slack.com How to Automate Slack
WebRun opens Slack to alert on idle GPUs. - Post an alert to your ML infrastructure channel naming every GPU instance that's been idle for more than 2 hours
- Include how long it's been running and its hourly cost
Done when The team has an alert only when a GPU is genuinely sitting idle.
How is each run configured?
Secure by default
Connect once, stays signed in
WebRun signs in once and keeps each session in a persistent environment, so every run picks up right where it left off.
Every action is checked against this policy before it runs.
Questions, answered
Will it stop or terminate an idle GPU instance on its own?
No. WebRun only reads utilization data and alerts you. Stopping or terminating an instance always stays a manual decision a person makes in CoreWeave.
What counts as idle?
An instance running with next to no meaningful GPU utilization for more than 2 hours straight, the pattern most likely to mean a job finished or crashed without anyone noticing.
Does it check every GPU on the account?
Yes. It reviews every GPU instance under the account each run, not just the ones tied to your main project.
Related templates
Put this on autopilot.
Turn it on in minutes - or have our team set it up for you.