Pay for what you use — compute per hour, agent work by token call. All prices USD.
Get StartedRun on demand or on a schedule to save.
Choose a region and machine when you build.
Billed only while your app runs.
| Machine | vCPU | Memory | GPU | Price / hr |
|---|
Machine rate × hours running.
Run on demand or on a schedule — you only pay while it's running.
Tokens are counted per prompt: $0–$40 each, up to 2 hours of autonomous AI agent work.
No seats, no tiers, no minimums.
Top up and go.