Start on Evaluation, prove value in one live workflow, and expand when your team needs shared controls.
Every plan supports the same delivery path. Expand reach and support as adoption grows.
Annual pricing is available where listed. Teams can still start on Evaluation without a credit card.
Best for individual evaluation and a first governed delivery workflow.
10 agents
Pilot one governed workflow with repo baseline checks for a single operator or repo.
Best for operators who need stronger enforcement and audit logs as usage grows.
100 agents· up to 5 seats
For operators who need stronger policy control, audit trails, and rollout visibility as AI-assisted delivery spreads.
Best for shared governance, review, and coordinated controls across engineering teams.
500 agents· up to 50 seats
For teams sharing policies, audit evidence, and rollout visibility across workspaces and repos.
Best for organizations that need custom rollout, security review, or commercial terms.
Unlimited agents
Custom rollout controls, proof workflows, identity and security review, and support for larger engineering organizations.
Evaluation is designed for real workflow proof, not a gated teaser. Paid plans expand reach, shared controls, and support.
Validate one repo with Evaluation or Pro — no procurement needed.
Move to Team when multiple operators need shared governance.
Enterprise adds rollout support, security review, and commercial terms.
The same product shape across every plan — repo checks, agent boundaries, CI, and shared visibility.
Higher tiers add reach and support, not a different control model.
Paid plans unlock shared governance and rollout support, not a separate product.
Yes. Start on Evaluation, then add shared governance and support as adoption grows.
No. Annual billing only changes price. The product surface stays identical.
When rollout touches multiple repos, operators, or sensitive enforcement needs.
No. Enterprise fits any team that needs custom terms or rollout support beyond self-serve.
If your team is already operating across multiple repos, multiple operators, or a compliance-sensitive environment, start with the pilot queue instead of guessing which plan fits.