Four plans, one meter. Pick the plan that fits how many agents you run.
CPU + memory, by the minute.
No compute. State kept.
Storage only, while you keep it.
Per-minute rates are published at launch.
hello@semantic.sh →
support@semantic.sh →