kibbu
Run AI inference on your own machines to cut cloud spend
Kibbu lets companies run AI inference workloads on machines they already own to lower costs and reduce reliance on cloud providers. It provides an admin console to add nodes, configure models, and control data routing, with zero data retention by default and options to keep sensitive workloads on-prem. You can start with a single workflow, swap in a Kibbu API key, and choose which models run locally while keeping cloud models where needed.
