Give a team GPUs without sharing a login
Everyone works under one billable account with their own membership and role, and every launch, stop and charge lands in one activity log.
Shared credentials are how a team ends up unable to say who started the run that is still billing.
Why it is hard
One login means no attribution
When everybody uses the same key, the audit trail records the account rather than the person, and every question about a charge becomes an argument about memory.
Removing someone means rotating everything
A shared secret cannot be revoked for one person. Somebody leaving turns into a change that interrupts everyone who stayed.
Spend is invisible until the invoice
A team that cannot see what it is spending as it spends it finds out at the end of the month, which is the one moment nothing can be done about it.
What answers it
Accounts own things, people belong to them
The billable client is the account. Each person is a separate member of it with their own role, so hardware and money hang off the team rather than off whoever happened to sign up.
Scoped keys, revoked one at a time
API keys are scoped and individually revocable. Removing one person's access does not touch anybody else's.
One log, everything in it
Launches, stops and charges land in the same activity log, in euro, as they happen rather than in a monthly summary.
How billing works
per second, to the microeuro · no minimum · no egress fee · idle costs nothing
Metering starts when the container does and stops when it stops. The rate is frozen the moment you match, so a provider re-pricing cannot change a running rental, and an aborted run costs what it used.
What you use from Kracht
Related jobs
- ComputeTrain a model without buying hardwareTake an H100 for the length of a training run and give it back. No reservation, no minimum, and no card sitting idle between experiments.Read the job
- ComputeFine-tune on a scheduleA job that runs nightly needs a GPU for the hour it runs, not for the day around it. Per-second metering makes the difference the whole business case.Read the job
- ComputeServe inference when traffic spikesAdd capacity for the hours you need it and stop paying when the spike passes. Nothing continues to bill after the container stops.Read the job