Baseten Production replacement. Truss-packaged model endpoints with autoscaling replicas — billed per prediction instead of the $1.80+ per-GPU-hour floor Baseten charges for always-on production tier. Built on MeterCall.
This module is your starting point. Describe what you want to layer on top — an interface, extra fields, a workflow, a whole app. Watch it build in real time. ⌘/Ctrl + Enter to run.
Your module's ready — tell us what you need
Use it, host it, give it a home, or keep building. You pick.