$49 per month
One person running an agent as part of real work.
- 3 agents
- 500 run credits
- Spend cap per agent
- Workspace kill switch
- Full audit log
Every plan carries the same controls. What changes is how many agents you run, how much work you can put through them, and how much of the reporting is done for you.
One person running an agent as part of real work.
Ops or engineering teams whose agents touch production systems.
Agencies, consultants, and MSPs running agents for many clients.
Many workspaces, custom residency, procurement, and SLA terms.
Teams and agencies that want the fleet standing correctly from day one. We stand the fleet up with you: isolation, spend caps set to your real budget, audit export wired into your review process, and the runbook your team keeps afterwards.
Your OpenClaw runtime and Claw, on your infrastructure, in your accounts.
Spend cap per agent, kill switch, approval gates on the work that deserves one.
Written for your team, not for us. Nothing here needs us to stay involved.
A run costs one credit, plus one more for every ten seconds it takes. That is the whole formula. Work that Claw refuses costs nothing, because it never reached the runtime.
Not per agent, not per seat. Adding a teammate or splitting one agent into three does not change the bill.
The room shows the cost of the last run. The credits page shows the ledger those numbers were added up from.
Each agent carries a ceiling. When it is reached, the run is refused and the refusal is recorded.