# Moonage Pricing

> Moonage is the organizational system for humans and agents. You pay a flat seat price
> for every human on the team. Agents do not take seats, and the models they run
> are passed through at the vendor's own cost.

| Item | Price |
|---|---|
| Human seat | $30/mo per human |
| Agent seat | Free — agents do not take seats |
| Model usage | Vendor cost, no markup |
| Compute | Metered per run |
| Billing | Prepaid wallet, $20 minimum, hard stop |
| Unused credit | Rolls over, fully spendable for 12 months |
| Plans | Core (self-serve) and Enterprise (custom) |

Canonical page: https://moonage.ai/pricing
This document: https://moonage.ai/pricing.md — also served from https://moonage.ai/pricing with `Accept: text/markdown`.

## How the pricing works

Two numbers, and the second one is at cost.

The first is the seat: $30 per month for each human on the team. Agents
are not seats — a team of four people can run forty agents and still pay four
seats. That fee is the business model.

The second is usage, drawn from a prepaid wallet. The tokens your agents run are
billed at the model vendor's own published price with no markup added, and every
line of every invoice shows that vendor cost beside what you were charged, so
the claim is checkable rather than promised. Compute is metered per run.

The wallet is prepaid with a hard stop, so spend cannot exceed what you have
loaded. The minimum load is $20, and credit you load rolls over and
stays fully spendable for 12 months.

## Plans

| Plan | Price | For |
|---|---|---|
| Core | $30/mo per human + metered usage | For teams putting agents on real work. |
| Enterprise | Custom pricing and terms | For organizations that need procurement, deployment, and control on their terms. |

Both plans are the same product. Every feature is included on both, with no
per-plan feature gates and no per-plan usage limits. Enterprise adds commercial
terms and compliance controls, not capability.

### Core

$30/mo per human + metered usage

- Full agent roster + supervisor coordination
- Managed models and every supported integration
- Model usage at cost — no markup
- Pooled wallet, per-member caps
- Governed organizational memory included

### Enterprise

Custom pricing and terms

- Volume rates on metered usage
- SSO / SAML, SOC 2, DPA
- Data residency / VPC options
- Postpaid invoicing, net-30, PO
- Named CSM, priority support

## What is metered

| Line | What it is | How it is metered |
|---|---|---|
| Models | The tokens your agents run | Which agent × which model, on post-cache tokens — at vendor cost |
| Compute | Sandbox and orchestration | Metered per run |

Nothing else is metered today.

## What is never billed

- **No surprise bills** — A prepaid wallet with a hard stop. Spend can’t exceed what you’ve loaded.
- **Failed runs never billed** — Runaway and errored runs are absorbed by us, never charged to you.
- **Credit rolls over** — Usage credit you load is fully spendable and stays valid for 12 months.
- **Models at cost** — Tokens pass through at the vendor’s own price — no markup, no keys to juggle.
- **Waiting is free** — reading, watching, idle time, and your own messages are never metered. You pay for what agents do.

## Example month

**These numbers are illustrative, not a quote.** They show the shape of a bill for a
four-person team, not what your team will pay.

| Line | Quantity | Billed |
|---|---|---|
| Human seats | 4 × $30 | $120.00 |
| Claude Sonnet 5 | 428 | $71.80 |
| Claude Haiku 4.5 | 1,264 | $14.60 |
| Compute | 16.6 hrs | $4.30 |
| **Total** |  | **$210.70** |

Against the same usage run unoptimized, that month saved $67.85 (43%) — $158.55 down to $90.70 on the metered lines. The drivers:

- Prompt caching — $21.40
- Smart model routing — $32.60
- Codemode — $13.85

## Core compared with Enterprise

### Billing

|  | Core | Enterprise |
|---|---|---|
| Seats | $30 / mo per human | Custom commitment |
| Billing flow | Seats + prepaid usage wallet | Postpaid, net-30, PO |
| Usage credit rolls over | 12 months, fully spendable | Custom terms |
| Per-member spend caps | Yes | Yes |
| Volume rates on usage | No | Committed-use discount |

### Agents & organizational system

|  | Core | Enterprise |
|---|---|---|
| Agents | Unlimited | Unlimited |
| Trust ladder — observe → advise → execute → autonomous | Yes | Yes |
| Multi-agent teams + shared memory | Yes | Yes |
| All models — Sonnet, Opus, Fable, Haiku, and more | Yes | Yes |
| Integrations & MCP tools | Yes | Yes |
| Supervisor coordination + delegation | Yes | Yes |

### Transparency & controls

|  | Core | Enterprise |
|---|---|---|
| Itemized invoice — vendor cost beside our price | Yes | Yes |
| Per-agent × per-model drill-down | Yes | Yes |
| Prepaid hard-stop + per-agent ceilings | Yes | Yes |
| Failed runs never billed | Yes | Yes |
| Budget alerts — 60 / 80 / 90% | Yes | Yes |
| Audit trail | Basic | Immutable, exportable |
| Governed organizational memory included | Yes | Custom retention |

### Security & support

|  | Core | Enterprise |
|---|---|---|
| SSO / SAML | No | Yes |
| SCIM / directory sync | No | Yes |
| SOC 2 + DPA | No | Yes |
| Data residency | No | Yes |
| White-label | No | Yes |
| SLA | No | 99.9% uptime |
| Support | Email | Named CSM, priority |

## Security and compliance

- SOC 2 Type II
- GDPR
- DPA
- SSO / SAML
- Data residency / VPC
- Encrypted in transit & at rest

## Questions

### What am I actually paying for?

A seat for each human on the team ($30/mo — agents don't take seats), plus usage: the tokens your agents run, itemized by which agent used which model, and the compute a run takes, metered per run. Reading, watching, idle time, and your own messages are free — you pay for what agents do, never for waiting.

### Do you mark up tokens? Why no bring-your-own keys?

No markup. Model usage is passed through at the vendor’s own price, and every line of every invoice shows that vendor cost beside what you were charged — so you can check it. The seat is the business model; the tokens are not. No BYOK keys to juggle either: the audit trail, governance, memory privacy, and trust ladder only hold up if the whole pipeline runs on our infrastructure under your rules.

### Can I get a surprise bill?

No. The wallet is prepaid with a hard-stop, every agent has a spending ceiling, and failed or runaway runs are never billed — we absorb them.

### What if I don’t use my whole wallet?

It rolls over, valid for 12 months. Credit you load is fully spendable until you use it — never a fee you lose.

### How is Core different from Enterprise?

Same engine underneath. Enterprise layers on a custom commitment, postpaid invoicing, SSO / SOC 2 / DPA, and named support — for teams that need procurement-friendly terms.

## Next steps

- Join the waitlist: https://moonage.ai/waitlist
- Contact the team: https://moonage.ai/contact

## More

- Product: https://moonage.ai/product
- Documentation: https://moonage.ai/docs
- Site overview for agents: https://moonage.ai/llms.txt
- Connecting an agent to Moonage: https://moonage.ai/auth.md