PRICING

You bring the keys.
Inference is at cost.

We charge for the platform: the runtime that keeps missions alive, the scheduling, the sandbox, the metering, the permissions. Your model provider bills you directly at their price. We take no cut of your tokens, which is unusual enough to say twice.

DECKHANDGetting a first crew running.
$0/ month
2 agents, 1 verified SSH hostScheduled missions, 15-minute floorOfficial plugins7-day transcript historyCommunity support
Start free
FLEETTeams sharing agents and hosts.
$24/ seat / month
Everything in QuartermasterSSO (SAML, OIDC) and SCIMShared account workspace and crewsPer-seat and per-agent permissionsQuotas and spend caps per memberAudit log, DPA, priority support
Talk to us
LIMITSDeckhandQuartermasterFleet
AgentsDeckhand2QuartermasterunlimitedFleetunlimited
Verified SSH hostsDeckhand1QuartermasterunlimitedFleetunlimited
Reverse tunnel hostsDeckhandQuartermasteryesFleetyes
Concurrent running sessionsDeckhand1Quartermaster12Fleet12 / seat
Schedule granularityDeckhand15 minQuartermaster1 minFleet1 min
HeartbeatDeckhandQuartermasteryesFleetyes
Community pluginsDeckhandQuartermasteryesFleetyes
Agent packsDeckhandQuartermasteryesFleetyes
Messaging channelsDeckhandQuartermasterTelegram, WhatsAppFleetTelegram, WhatsApp
Workspace storageDeckhand1 GBQuartermaster50 GBFleet50 GB / seat
Transcript historyDeckhand7 daysQuartermasterunlimitedFleetunlimited
Spend metering & quotasDeckhandsession onlyQuartermasterfullFleetfull + per-member caps
SSO / SCIMDeckhandQuartermasterFleetyes
Audit logDeckhandQuartermasterFleetyes
SupportDeckhandcommunityQuartermasteremailFleetpriority + DPA
WHY THIS SHAPE

Bring your own keys, and what it saves you.

Platforms that resell inference add a margin to every token, which is invisible until your usage grows and then becomes the whole bill. A crew of agents running long missions is a lot of tokens. You should pay your provider's rate for those, and pay us for the part we actually built.

Your provider, your rate
Register any provider with your key, add its models with per-token pricing, and point each agent at the model that suits its job. Cheap model for triage, expensive one for architecture.
Metered, not guessed
Spend recorded per session, per agent and per model, with quotas that stop a runaway mission before it becomes an invoice.
No lock on the way out
Your agent configuration, transcripts and workspace files export as plain files. We would rather you stay because it works.
QUESTIONS

Common questions, answered plainly.

Do you charge for tokens?No. Your provider bills you at their price. We never sit in that path and take no percentage.
What counts as a seat?A person with their own login on a Fleet account. Agents are not seats — you can run as many as you like.
Do subagents cost extra?Not on our side. They consume tokens from your provider like any other session, and their spend is metered separately so you can see it.
Can I self-host this?No. If self-hosting is a requirement, OpenClaw is a genuinely good answer and we say so on the comparison page.
What happens if I stop paying?Missions stop firing and agents go idle. Nothing is deleted for 60 days, and your export stays available the whole time.

Crews of expert agents that finish the work.

Create an account, add your model key, and write your first agent's soul. The rest of the crew can arrive in one click.