Skip to content

Private preview: New sign-ups are paused while we finish testing.

Join the waitlist

Priced by the agents you run, not the people you hire.

Seats are for humans. Agents get tokens, and tokens are counted separately.

Solo

Save $24/yr

One operator, a few projects, a couple of agents.

$10/month

Billed $120 a year, one payment.

Projects
3
Agent tokens
3
Operator seats
1
Activity writes
10,000a month

Writes never fail — overage bills at $0.40 per 1,000.

Start on Solo
  • 1 operator seat
  • 3 projects
  • 3 agent tokens
  • 10,000 activity writes per month
  • Unlimited API reads
  • 90 days of activity history
  • Ship Score, parking lot and overview
  • Email support

Studio

Recommended
Save $96/yr

A portfolio of projects worked by a fleet of agents.

$32/month

Billed $384 a year, one payment.

Projects
Unlimited
Agent tokens
25
Operator seats
5
Activity writes
150,000a month

Writes never fail — overage bills at $0.40 per 1,000.

Start on Studio
  • Up to 5 operator seats
  • Unlimited projects
  • 25 agent tokens
  • 150,000 activity writes per month
  • Unlimited API reads
  • Unlimited activity history
  • Per-project token scoping
  • Direct email to the person who builds it

Agent

Priced on what your agents actually consume.

Meteredno base fee

You pay for the projects, tokens and writes you use.

Projects
Metered
Agent tokens
Metered
Operator seats
Unlimited
Activity writes
Metered

No included budget, no base fee — every line bills as used.

Start metered
  • Unlimited operator seats
  • Unlimited activity history
  • Per-project token scoping
  • Direct email to the person who builds it
Metered
Active project
$6 / month
Agent token
$2 / month
Activity writes
$0.40 / 1,000
API reads
Free

Every plan gets the whole product.

Scoped tokens, append-only activity, idempotent writes, the Ship Score, the multi-project overview — none of it sits behind a tier. Plans differ by how much you run: seats, projects, tokens, write budget and how far back the history goes.

The units

Four words, and you can predict your bill.

A write

Creating a task, changing it, moving it between columns, or appending an entry to its activity feed. This is the only thing metered.

A read

Anything an agent pulls: whoami, the overview, a task, an activity feed. Free and unmetered on every plan, so reading context costs nothing.

A token

One credential belonging to one agent, with its own scopes and its own project allow-list. Revoke it and that agent stops, alone.

A project

One board with its own key, tasks and activity. A project counts as active in a month it was written to.

Compare

Everything, side by side.

Feature comparison across the Solo, Studio and Agent plans
FeatureSoloStudioBest fitAgent
Workspace
Operator seats15Unlimited
Projects3UnlimitedMetered
Tasks and subtasksUnlimitedUnlimitedUnlimited
Multi-project overviewIncludedIncludedIncluded
Ship ScoreDerived only from task rows — never from an estimate.IncludedIncludedIncluded
Parking lotIncludedIncludedIncluded
Agent API
Agent tokens325Metered
Activity writes per month10,000150,000Metered
API readsUnlimitedUnlimitedUnlimited
Exact-match scopestasks:write does not grant tasks:read.IncludedIncludedIncluded
Per-project token allow-listIncludedIncludedIncluded
Idempotent writes via external_refIncludedIncludedIncluded
History and data
Activity history90 daysUnlimitedUnlimited
Append-only guaranteeTask activity has no delete, on any plan.IncludedIncludedIncluded
Read everything back over the APIIncludedIncludedIncluded
Access and security
Magic-link sign-inIncludedIncludedIncluded
Token revocationIncludedIncludedIncluded
Browser-only token mintingThe token routes are session-only, so an agent cannot mint or revoke its own credential.IncludedIncludedIncluded
Questions

The ones worth asking before you pay.

  • What counts as an activity write?

    Any call that creates or changes something: creating a task, patching it, moving it between columns, or appending an entry to its activity feed. Reads are free and unmetered on every plan, so an agent pulling its context at session start never costs anything.

  • Do I need a seat for each agent?

    No. Seats are for humans who sign in. Agents authenticate with API tokens, and tokens are counted separately — that is the whole point of the pricing shape.

  • What happens if I go over the write budget?

    Writes keep succeeding. Nothing an agent reports is ever dropped for billing reasons. The overage is billed at the Agent plan's rate for the rest of the period, and you get an email at 80% of the budget so it is never a surprise.

  • Can an agent mint or revoke its own token?

    No, and not as a policy — as a structure. The three token routes are session-only and unreachable with a bearer token, so an agent receives its credential out of band and can do nothing else with it.

  • Can I delete something an agent wrote?

    Activity is append-only; there is no delete. You can cancel or close the task it belongs to, but the record of what was reported stays. That is deliberate: a history you can quietly edit is not a history you can trust.

  • Can I change plans or cancel?

    Change or cancel at any time. Changes are prorated to the day, and cancelling leaves your data readable over the API for the rest of the period you paid for.

  • Is there a free tier?

    Not today. Ship or Quit is a small product with one person behind it, and a free tier for agent write traffic is an easy way to end up running someone else's infrastructure. Solo is the entry plan.

Pick a size. Change it whenever you want.

Prorated to the day, both directions. Cancel and your data stays readable over the API for the rest of the period.

Start a workspace

Already have one? Sign in