Pricing
Safety is never the upsell.
Every plan runs the same governance core — the policy engine, approvals, redaction, and audit. You pay for capacity, not for the right to be safe.
Failed to fetch
Questions, answered
Is governance an add-on?
No. The policy engine, approval inbox, column redaction, and audit trail are the platform — every plan runs the full governance core. Plans differ on capacity and scale, never on safety.
Do I need to bring my own AI model?
Yes — connect your own LLM provider keys (Anthropic, OpenAI, and compatible providers). Your agents run on your model under your policies, and we never resell your tokens.
What happens when an agent tries something risky?
The action pauses before it executes and lands in your approval inbox. A human approves or denies it; on approval the run resumes exactly where it stopped. Denied actions never touch your data.
Can I start free?
Yes. New accounts start on the free plan with no card required, and you can upgrade from the console when you need more capacity.