A flat software fee.Inference is billed separately.
Two decisions, priced separately: how you run the software, and where inference comes from. There is no per-seat line anywhere on this page.
No surprises
Spend stops at a cap.
Every agent and every run has a limit. When an agent reaches it, it stops, and the bill stays where you set it.
- A cap per agent
- Set how much one agent, and one run, may spend.
- Itemized inference
- Model use is its own line, so you can see what each agent cost.
- No per-seat charge
- The software fee is flat. Adding a person adds no line.
Spent so far
$0.50
The agent stopped at $0.50. Nothing more is spent until a person raises the cap.
Pick how you run it.
Self-hosted license
For teams that run their own servers. You license the software and run it on your own hardware, with your own database. It keeps running there for as long as your license stands. Fee: a flat software fee.
Talk to usManaged
For teams that should not run servers. The same software and the same data-ownership posture, operated by us on infrastructure you authorize. Fee: a flat fee per workspace per month.
Talk to usDelivery service
For teams that want the agents built for them. We build and run agents for your operation, inside an APG (Agents Proving Ground) you control. Fee: a setup fee plus a monthly retainer, scoped to your operation.
Pick where inference comes from.
Your own model key
Your model provider bills you directly and we take no inference margin. Fits teams with an existing model contract.
Your own GPU
A local model on your hardware. There is no inference bill from us or from a model provider. You pay for the hardware and the electricity. Fits the highest-privacy work.
Metered through us
Billed per token and itemized on your invoice. When it opens, we state the per-token markup before you sign.
What a quote contains.
The software fee for the products you license, the managed fee if you want us to operate it, the delivery setup fee and retainer if you want agents built, and your inference tier. We confirm each figure after a short scoping conversation, because the numbers depend on scope.
What happens after you ask.
A named person reads your message and replies within 48 business hours. They ask what the work is and tell you whether it fits. There is no marketing automation behind the form.
Detail view: the pricing matrix
Read across for the inference tier, down for how the software runs.
Inference · metered metered, itemized | Inference · your own key your model key | Inference · your own GPU your own GPU | |
|---|---|---|---|
Self-hosted Flat software fee | Flat fee · metered inferenceComing soon. Licensed software on your servers, your database, with inference billed per token through us and itemized on the invoice once this tier opens. Until then, start on your own key. | Flat fee · your model billLicensed software on your servers. Your own key, so your model provider bills you directly and we take no inference margin. Fits teams with an existing model contract. | Flat fee · your own GPULicensed software and a local model, both on your hardware. Nothing leaves the building and there is no inference line from us or from a model provider. Fits the highest-privacy verticals. |
Managed Flat / workspace / mo | Flat fee · metered inferenceComing soon. We operate it on infrastructure you authorize, with inference billed per token through us and itemized once this tier opens. Until then, managed operation runs on your own key. | Flat fee · your model billWe operate it; your own key handles inference. Fits managed operations with an existing model contract. | Flat fee · your own GPUWe operate it against your own GPU and local model. Fits a sovereignty mandate that still wants managed operations. |
We confirm software fees per engagement, because the number depends on scope. Metered inference through us is coming soon. When it opens we will state the per-token markup before you sign. Here is how inference is itemized.
Agents and apps.
One flat monthly total, shown before you deploy, with the saving from bundles worked out for you. There are no credits and no per-use meter. Every purchase has a 30-day refund. Agents are installed for you by a representative. The apps are not open yet, so they cannot be bought today.
For a business
USD 5.99 to 9.99
Per agent, per month. Steady jobs such as a log or a folder in order are 5.99. Most other jobs are 9.99. Pay for a year and you pay for ten months.
USD 49 a month
Premium: nine agents for one workspace, with agents that hand work to each other and approval before anything goes out. Bought one by one the same nine come to USD 92.91 a month, so Premium saves 47%.
USD 79 a month
Premium plus Case Room’s four add-on agents, which cost USD 128.87 a month separately. That saves 39%.
For you, personally
Ten single-purpose apps that run on your own computer, paid once, each about USD 10 to 60. Any three for USD 59.99. All ten for USD 179.99. They are for personal use, not business data.
- Bill and subscription watchdog, about USD 19.99
- Home paperwork, receipts and warranties, about USD 24.99
- Sole-trader books and invoicing, about USD 59.99
- Contract fine-print reader, about USD 14.99
- Data-broker removal, about USD 29.99
- Photo organizer and backup, about USD 29.99
- Personal health record, about USD 29.99
- Car and home maintenance log, about USD 9.99
- Small landlord rent and tenant manager, about USD 49.99
- Digital legacy folder, about USD 39.99
Case Room
USD 49.99, paid once. Add-on agents are USD 5.99 each, or all ten for USD 39.99. You pay your own AI provider separately if you bring your own key.
Why the pricing looks like this.
Why no per-seat pricing
Per-seat pricing taxes you for growing. It ties your bill to headcount and ignores the work the software does, and it charges you for the colleague who barely logs in. Our costs do not rise when you hire, so our price does not either.
There is a second reason now. When agents do the work, a seat count has stopped describing anything real. You are buying a department that runs, and a login count says nothing about it.
Why inference is itemized
Bundling inference into a plan hides what it costs and lets the margin drift. We will buy the capacity, meter what you draw and show our margin as a line on the bill, and let you decide whether to buy through us or bring your own key. A bill you can read is a bill you can leave.
Why the managed tier exists
Some teams should not run servers. The managed tier is for them: the same software, the same data-ownership posture, operated by us against infrastructure you authorize. It is a convenience you pay a flat fee for, and we do not use it to charge you more as you grow.
What you are licensing
The software is proprietary and licensed to you. You get the runtime and you run it on your own hardware, and it keeps running there for as long as your license stands, whether or not you renew a retainer with us. You do not get the right to redistribute it.
What about having it built for you
That is our delivery service: a setup fee plus a monthly retainer, scoped to your operation. The case room, the brain and the personal assistant have no list price. We quote them as a delivery engagement, or as a product once one ships.
How the delivery service works →Talk through the fees and pick an inference tier.
When you want the numbers for the license, managed operation or metered inference, a call is the fastest path.