Pricing
Every plan runs the whole system — state, strategy, missions, verification. Plans set capacity; you set how far it goes.
The real loop, once, for real.
$0
100 credits / month
3 seats
Real limits, honestly drawn — enough to know if this is your new way of working. The Free plan IS the trial — no clock, no card.
Start freeFor a company in motion.
$29/mo
1,000 credits / month
7 seats (+$7/seat/mo)
For a team operating daily.
$99/mo
4,500 credits / month
20 seats (+$7/seat/mo)
For companies running serious delivery through the Fleet.
$299/mo
16,000 credits / month
50 seats (+$5/seat/mo)
The autonomy ladder is the premium lever — ceilings L0–L3 · L0–L5 · L0–L6 · L0–L7 across the four plans: the more your company lets Brainis run, the more the higher tier pays for itself.
Your subscription is access: seats, products, limits, and a monthly allowance of AI credits — the variable capacity your strategy runs, agent work, verification, and predictions consume. When you need more, top-ups add credits to the workspace without changing your plan. Every credit-consuming action shows its cost before it runs.
Every metered action carries a named class, and the class caps what it can cost: assist (Cortex questions) · brief — your daily brief, text or audio, its own class so a brief can never bill at report-band · report (deliverables) · autopilot (agent work) · scenario (predictions) · analysis (verification). One currency, one balance: voice folds into the same credit display (1 min studio voice = 2 credits) — one number, one word, never a second meter.
Live price flow
Before you run it
Top-up packs
| Pack | Price | Credits |
|---|---|---|
| Boost | $10 | 400 |
| Sprint | $25 | 1,100 |
| Launch | $75 | 3,600 |
| Scale | $200 | 10,500 |
| Fleet | $500 | 28,000 |
Included monthly credits reset each cycle (no rollover); purchased top-ups roll for 12 months, consumed after the monthly grant. Workspace-owned; billing roles buy, anyone else requests via the decision queue. Pack $/credit is deliberately slightly worse than upgrading a tier at equivalent volume — bursts stay friction-free, sustained heavy use converts to the next tier honestly.
| Action | Credits | Class |
|---|---|---|
| Cortex question | shown in-product before every run | assist |
| Daily brief (text or audio) | shown in-product before every run | brief |
| Report / deliverable generation | shown in-product before every run | report |
| Agent work item (draft, research, build step) | shown in-product before every run | autopilot |
| Scenario / prediction run | shown in-product before every run | scenario |
| Verification pass | shown in-product before every run | analysis |
| Strategy Engine full blueprint (A to Z) | shown in-product before every run | composite — Σ of per-step charges |
| Typical full mission (e.g., revive stalled deals) | shown in-product before every run | composite — Σ of per-step charges |
| Non-AI product use (CRM edits, boards, docs, dashboards) | 0 — never metered | |
Every credit-consuming action shows its exact cost before it runs. Published bands land here once real per-action telemetry confirms them — never before.
Need a custom credit pool or on-prem deployment? Talk to sales → about Enterprise.
No. There is no implementation package, no managed service, no consultants — software all the way down. Design partners get early access and roadmap influence, not our hands on their work. If the AI can't do something, refer it to us: we build the capability.
Included monthly credits reset each cycle; purchased top-ups roll for 12 months. The exact policy is in plain text at purchase, not discovered later.
The Free plan IS the trial — the real loop with real limits, no clock, no card. One exception: a 14-day Growth trial, no card, offered only when you actually hit a paid edge (an upgrade-consideration moment), never at signup — one per organization, ever. When it ends you revert to Free in one transaction: data intact, purchased top-ups persist, and anything paused resumes when credits do. Trial expiry is not dunning — there is no card, so nothing retries.
Billing roles. Anyone else can request capacity through the app's decision queue — the request lands with the person who can approve it.
Only through an authority contract with an explicit budget and, for irreversible spend, a human gate. External spend is never paid in AI credits — the two are separate on purpose.
We retry on days 3, 5, and 7, with plain notices each time. At day 7: grace mode — reading and decision approvals stay, agent execution pauses (running missions checkpoint, never killed mid-step). At day 21: downgrade to Free-plan capacity, data intact per the retention policy. Payment is the un-pause — capacity restores and paused missions resume from checkpoint in the same moment.
30-day money-back on your first subscription charge, no interrogation. If an incident on our side touches a mission, its credits refund automatically — we never charge for our failures. Top-ups are non-refundable once more than 20% consumed; that line is shown at purchase, not discovered later.
Export everything — our Terms give you a 30-day window after cancellation to get your data out, after which it may be deleted per the DPA. Leave-anytime is a design goal, not a retention trick.
Once, genuinely: state → strategy → one governed, verified mission. Limits are on quantity and seats — not on truth of the experience.