Local chat
Ask, draft and reason with a model that runs on your machine. The conversation is produced locally, and it is not billed by the token.
Applications
One private runtime. Several ways to use it.
Every application below runs on the same hardware plan, the same verified model profile and the same local boundary. Each is listed because the product data records it as working, with its standing beside it.
Local API for developers
An OpenAI-compatible endpoint on loopback, authenticated with a generated key, for your own scripts and for existing clients.
Private documents
Grant the folders you choose, ask in ordinary language and check each answer against the cited file and page. When the folders do not cover the question, it declines.
Local chat
A private assistant that answers on your machine.
Chat uses the model profile recommended for your hardware. The prompt and the answer stay on the machine during local inference, and use is not billed by the token. Anything that speaks the OpenAI-compatible API can also serve as the chat front end.
Professional deployment
For organisations, a managed deployment.
Organisations that want AI Stays Local on workstations they own can agree an Enterprise Managed arrangement: deployment planning, hardware validation for the machines they intend to use, model planning from measured evidence, and support. It is a custom commercial agreement, not a plan with a price list.