Plans and billing | kanman documentation

kanman is billed per AI team, not per seat. Every member of your workspace can work with every team at no extra cost.

Current prices are on the kanman pricing page. This page explains how the plans work.

Plans

Plan	For	Runs at a time per team
Trial (no plan)	Every new team, until you choose a plan	1
Pilot	A fixed-fee, six-week pilot for one team	up to 3
Team	One AI team in production, billed monthly	up to 5
Self-hosted	Teams whose runs execute on kanman’s runner inside your own network, on request	up to 20

The number of runs at a time is a cap. Your team’s policy decides how many runs it actually starts (max_concurrency); the effective value is the lower of the two. See Policy settings.

Each team in a workspace has its own plan. A workspace with three teams can have one on a pilot, one on the Team plan and one still in the trial.

A pilot ends by itself six weeks after it starts. After that, or when a Team plan ends, the team goes back to the trial: it keeps working, one run at a time.

Compute: your key or pass-through

The coding agent and the reviewer use language models. Each team pays for that compute in one of two ways:

Billing mode	How it works
Pay through kanman (pass-through)	kanman provides the model access and bills the compute at provider cost plus 15 percent, as a separate line on the team’s invoice. No key needed.
Use your own key (bring your own key)	Runs use your workspace’s own Anthropic, OpenAI, Azure OpenAI or Amazon Bedrock key. Your provider bills the model costs to you directly.

You choose the billing mode in the Executor step when you set up a team. The team settings show it under Budgets. The workspace key itself is managed in Settings > AI providers, see What model providers see.

Pass-through compute is invoiced for teams with a plan. Whichever mode you use, your team’s budgets apply: per run, per day and per month. kanman stops before a budget is exceeded and asks in the decision inbox. See Budgets and cost.

Managing billing

Workspace owners and admins manage billing.

Open Settings > Billing (/<workspace>/settings/billing). The command palette entry Plan & billing (Ctrl+Shift+S, Cmd+Shift+S on Mac) opens the same page.
The page shows the three plans and, under AI teams, each team with its plan, how many runs it may start at a time, and when the plan renews or the pilot ends.
For a team in the trial, choose Start Team plan or Start pilot. You complete the payment in a secure checkout and return to the billing page. The plan is active as soon as the payment is confirmed.
For the Self-hosted plan, choose Ask about self-hosted and we set it up with you.

Invoices and payment details

Choose Open billing portal to see and download invoices and to update the payment method, billing address and VAT ID. All teams of a workspace share one billing account, so you get one invoice stream and enter your VAT ID once. The portal is available after the first plan has been started.

The cost of each run is shown on its run page and summed up in the weekly team report.

Ending a plan

End a team’s plan in the billing portal. It ends at the end of the current billing period. The team then goes back to the trial with one run at a time. Its history (runs, evidence, decisions) stays readable.

If a payment fails, the team shows Payment due and keeps working while the payment is retried.

Questions

For pilots, the Self-hosted plan or invoices, write to [email protected].