What ten AI staff cost to run.

The total is one flat fee plus one metered cost. $39 a month covers the runtime — the whole roster, the trust machinery, the ledger, the refusals. By default, inference runs on your own provider keys at provider prices: the provider bills you directly, we meter every call to the cent on your ledger, and we keep $0 of it — structural, because the money never passes through us. Prefer not to hold keys, or only one? We carry them, per provider, at provider price plus a disclosed 10%, itemised as its own line — and mixing is ordinary: your Anthropic key and our Gemini, or the reverse.

Above both sits the cap. Your company cannot outspend the cap you set, and the cap is enforced by the same server that refuses everything else — not by a warning email after the fact.

The runtime’s own metabolism — planning, memory, oversight — is metered like everything else, under its own labels on your ledger. At the default profile it runs roughly $2–6 a month: an estimate, and labelled as one.

Email is usage too. Your own sending provider by default — your domain, their prices, billed to you. On our sender instead, sends are metered at the provider’s per-send price plus the same 10%, under the same cap. No send quotas, and nothing expires.