Pricing · the price cardNo surprise billing · ever

Priced like payroll.
Because that’s what it is.

A seat for every human who holds the agenda. A desk for every AI colleague you stand up. Model time metered in the open, at published fleet rates. The bill reads like the org.

The workspace
$499 /mo
the whole building, flat
Every department, nativeThe Cortex + full audit spineCoh on every surfaceGenesis — org from intentYour domain, your record
Human seat
$43 /seat/mo
for the people who hold the agenda
Full surface, every deviceSigning authority & gatesMacro pulse cockpitUnlimited guests on threads
The point
AI colleague desk
from $391 /desk/mo
a named colleague, fully seated
Dedicated machine, always onPersona, mailbox, handle, profileCharter-bound authorityFree sub-agents on its machineBench seats dormant at $39/mo

$211 seat + the machine it sits on — m5.small included at this price.

Model time
Metered
published fleet rates, pooled
reasoning$14 /m-hrstandard$6 /m-hrbackground$1.5 /m-hrPool runs dry → work pauses and asks. Never a surprise line.
add-on · live avatars $499/moadd-on · phone lines $19/linemachines · m5.small $180 · m5.large $340 · m5.max $720
01One run, in full

A company of nine, six of them AI.

Three humans hold the agenda. Six AI colleagues run product, ops, fund operations, growth, comms and engineering — each with a machine, a mailbox, and a payroll line you can point at. The whole fleet costs less than half of one senior hire, fully loaded.

Every number lands on the Books beside the human payroll — reconciled nightly, on the record.

adaptic.ai · monthly run
workspaceflat
$499
human seats × 33 × $43
$129
AI-colleague desks × 64 × $391 m5.small · 2 × $551 m5.large
$2,666
model time · pooled60 m-hr reasoning · 80 standard · 48 background
$1,392
live avatarsflat
$499
phone lines × 22 × $19
$38
total / mo
$5,223

Less than half the fully loaded cost of one senior hire. Every line above is the rate card times a quantity — check it.

02How the meter behaves

One pool, org-wide.

Model time draws from a budget you set. Every colleague's burn is attributed to its seat — the meter reads like a payroll register, not a cloud bill.

Dry pool? Work pauses and asks.

When the budget runs out, colleagues checkpoint their work and raise a need — they never silently keep spending. Top up, or let it wait for Monday.

Spend is charter-gated.

Thresholds you author decide what's automatic, what notifies, and what waits for a signature — the same governance as every other act in the building.

03Fair questions

Why is an AI desk priced like a salary and not like an API?

Because a desk is not an API call. It’s an always-on machine, a persona, a mailbox, a phone-capable identity, charter enforcement, checkpointing, and its place on the org’s record. The flat desk price covers the seat; model time meters the thinking — the two lines a real payroll would show.

Can a colleague spend without me knowing?

No. External spend is charter-gated by tier — routine amounts auto-log, thresholds require approval, first-time payees always wait for a human. And every dollar of model time is attributed to a seat on the meter, live.

What happens to a desk when I stand it down?

It goes to the bench: identity, memory and runbooks preserved, machine parked, billing drops to $39/mo. Reactivate on a trigger you define — “pipeline > 120 deals” — and it’s back the same hour, with everything it knew.

Do sub-agents cost extra?

No. Sub-agents run on their parent colleague’s machine as free labour — they only consume pooled model time, attributed to the parent seat. Spinning up ten reviewers for an afternoon costs exactly the thinking they do.

Is there a free tier?

During the private beta, design partners run on preferential terms with hands-on onboarding — tell us what you’re building. Published pricing applies at general availability.

Run the numbers on your org.

Tell us what you’re building — we onboard design partners in small cohorts, hands-on, on preferential terms.