Enterprise
The engine is open source and free. Companies pay us to set it up and run it on their own machines, and for Lumbridge 1 by the token.
Open sourceApache-2.0 · no license fee
12 parts you run yourself
From usPaid · each on its own
Three things we do
01 · Work with us
Two services, each with a person at our end who answers for it. The engine runs on machines you own; we put it there, get your first agents working, and keep it running with you.
We install the engine on your machines and get your first agents running.
Quoted in writingOne fixed price, quoted in writing after one call
Ongoing, with someone accountable.
Quoted in writingMonthly, quoted in writing with the setup
02 · How we work
You tell us the team, the machines, and what you’d hand to agents first.
Scope, price and response time, written down before any work starts.
We install Lumbridge, wire it to your code host and mail, and set up the Factory on your repositories.
You get a runbook. Support continues with a named engineer if you want it.
03 · What your team sees
Both run on a laptop with nothing from us in the loop.
Gates
An agent that needs an answer says so and stops. Sending mail, merging code and spending money stay with people.
Layouts
Panes resize, move and split. A layout is a file, so a team can share one or each person can keep their own.
04 · How it fits a company
Laptops, a build server, a GPU box, rented servers. Nothing needs our cloud to work.
Agents draft and build. Sending mail, merging code and spending money wait for a person.
Any OpenAI-compatible endpoint works: Lumbridge 1, or open models on your own hardware with Lumbridge Computer.
Apache-2.0. If we disappear, you still have the source and your data in plain files and standard databases.
05 · Straight answers
No. We are a small lab and hold no certifications today. We answer security questionnaires in writing and can work under your agreement.
Not yet. Core gives every client on a machine the same access, and the Library has one shared token. Both are on the list; we’ll tell you where they stand.
We host the model API. We don’t host Lumbridge itself: it is meant to run on machines you control, and we help you set that up.
Small. You talk to the engineers who write the code, and we take on a limited number of companies at a time so that stays true.
The model is priced by the token on the Models page. Setup and Support are quoted in writing after one call.
Tell us how many people, which machines, and what you’d hand to agents first.