Pay for work, not seats at a chatbot.
Plans are sized by the work agents actually do, measured in credits. Budgets are enforced before spend, on every plan.
Free
The trial. One agent, supervised, on the record.
- 1,000 credits / month
- 1 agent
- Supervised trust level
- 7-day audit retention
- 1 seat
Starter
Your first real delegation. Approvals on, budget capped.
- 25,000 credits / month
- 3 agents
- Supervised + Approve-sends
- 90-day audit retention
- 3 seats
- Credit top-ups
Pro
Most capableA full roster, at every trust level your agents have earned.
- 150,000 credits / month
- Unlimited agents
- All trust levels, including Autonomous
- 1-year audit retention
- 10 seats
- Credit top-ups
Enterprise
Custom credit grants, custom terms, and a direct line for your security review.
Early pricing, subject to change.
One unit. Fully legible.
1 credit ≈ a tenth of a cent of model work
Cheap models cost few credits, premium models cost more. The meter is visible in your workspace, per agent, per run.
Budget caps stop spend before it happens, not after
Caps are real dollars, set per agent. The Budget Gate blocks any call that would cross the cap, then tells you why.
Bring your own API keys and skip credits entirely
Model usage bills to your provider directly, on your terms. Caps and the audit log still apply.
The fine print, plainly.
How credits, caps, and billing actually behave. No surprises by design.
What is a credit?
What happens when my credits run out?
Do unused credits roll over?
Can I buy more credits mid-month?
How do budget caps relate to credits?
Can I skip credits entirely?
Is there a free trial?
Hire your first agent.
Keep the final say.
One agent, one job, one budget. Free tier, no card. Every action on the record from the first run.
Hire your first agent