Your models, your machines. Nobody else’s.
Connect each machine as a node, pin a key to “My nodes only,” and inference runs only on hardware you operate — fail-closed, at zero platform fee.
On your hardware only
Pin a key to “My nodes only” and requests route exclusively to nodes you operate. If none can serve, the request fails — never a silent spill to someone else’s machine.
Your models, zero platform fee
Your own models run on your own machines at no platform fee. We handle routing, scheduling, and uptime so your team ships against one OpenAI-compatible API.
Self-host or air-gap
Run the whole platform — the Hub included — on your own infrastructure, air-gapped if you need it. Nothing leaves your network.
Setup is two steps: connect each machine as a node — the same install you’d use to run a node — then pin an API key to “My nodes only.”