A model partnership built around access
Capix is the official compute partner of the independent SuperGemma Foundation. Together, the projects are making a curated set of open-weight models easier to discover, quote and deploy through the same Capix balance and control plane used for other compute.
The partnership is practical: SuperGemma brings community, model curation and ecosystem coordination; Capix provides the deployment path, private endpoint, usage metering and access from developer tools.
From model page to private endpoint
A customer selects a featured model, receives a live capacity quote and confirms the deployment. Capix then allocates compatible GPU capacity, starts the serving runtime, downloads the immutable model revision and waits for health checks before marking the endpoint ready.
Once healthy, the endpoint belongs to that customer. It can appear in Capix Chat, Capix IDE and Capix Code without becoming a public model for other accounts.