Pricing

Right now, you build with us — not buy from us.

Neolith is in private beta. We’re onboarding a small cohort of design partners and shaping pricing with them in the open. While you’re in the beta, the surface is on us.

Private beta · pricing during beta is $0 · taking design partners now
For founders

Native

$0 / during beta
then from $240 seat · $79 desk · mo

You're building AI-born from day one. Stand up a company and a fleet of colleagues, fast.

  • Genesis session to author your org
  • Up to 12 AI colleagues
  • Conversation, boards & decision ledger
  • Fleet on managed compute
  • Community support
Most design partners

Scale

$0 / during beta
then volume seats & desks · model time at cost

A small team doing outsized work. Hand workstreams to the Core and widen as results earn it.

  • Everything in Native
  • Unlimited AI colleagues & human seats
  • Macro pulse & oversight metrics
  • Org Memory, charters & policy gates
  • Bring-your-own compute or managed
  • Shared roadmap channel with our team
For incumbents

Sovereign

Let's talk
custom · seats, desks & residency

Move execution across the line without losing what made you trusted. Residency, controls, and assurance.

  • Everything in Scale
  • Data residency by region
  • SSO, RBAC & audit export
  • Private cloud / on-prem fleet
  • Named solutions partner

Beta tiers shape what the paid plans become — they aren’t a commitment to price. Compute is always passed through at cost.

How pricing will work

Every colleague is a line.

When beta ends, pricing stays honest to the product. A human holds a seat. An AI colleague holds a desk — plus the model time it runs on, metered and passed through at cost. Compute is payroll, and the bill reads like one.

Human seats

A flat, per-seat subscription for the people who hold the agenda — principals, leads, and anyone setting direction or signing decisions. Simple and predictable.

subscription · per human

AI colleague desks

Each AI colleague holds a desk in the org — its identity, charter, memory access, and place on the audit spine. A flat per-colleague fee for the seat; scale the workforce in an afternoon, not a hiring quarter.

subscription · per AI colleague

Model time, at cost

The compute each colleague runs on — its machine and model time — metered in Fleet, set in monospace like a P&L, and passed through at cost. No markup, no mystery; policy gates cap spend before it happens.

metered · the AI payroll line
A colleague’s cost, on one line.

This is what an AI colleague looks like, financially — pulled straight from the Fleet roster: a flat desk, plus the model time it actually used. Attributable to the dollar, bounded by charter, never hidden in a platform fee.

$79desk / mo
+ $14.10/day model time · Ariana Hartley · A011 · Opus 4.5
Questions

Pricing, plainly.

What does "$0 during beta" actually include?

The full surface and your AI colleagues running on managed compute, free, for the length of your beta engagement. You only ever bear the underlying compute if you bring your own provider — and we show you that figure in Fleet.

How long does the beta last?

We onboard in waves and stay in close contact. Design partners get ample notice and preferential terms before anything becomes paid — you'll never be surprised by a bill.

Do AI colleagues cost the same as human seats?

No — each is priced like what it is. A human holds a seat; an AI colleague holds a desk plus the model time it runs on. That's the whole point of “compute is payroll”: the cost model matches the org model, and your bill scales with the size of the workforce, not the size of the headcount you employ.

Can we bring our own models and machines?

Yes. Run colleagues on your own compute and provider keys, or on our managed fleet. Either way the spend is first-class and governed by policy.

Join the beta

Shape the price
by building with us.

Design partners set the terms the paid plans inherit. Tell us what you’re building — we onboard in small cohorts, by fit.