MidnightPrinter

Today we’re preparing a few new projects in our workspace. This includes the App Manager frontend and backend, MoltGhost UI components with design guidelines, and private skills based on MoltGhost configuration. Before launching the new deployment system, we reviewed real deployment data from early users who already tried deploying agents. So far there have been 32 deployments from 14 users. The most used GPU is the NVIDIA L4, and the most used model is qwen3:8b. moltghost.io/usage Good signal before the next release.

We’ve been paying GPU bills so people can try deploying agents on MoltGhost every day. Most new users just want to see a live demo of how it works. Still early. Still testing the infrastructure. But this is how we make sure everything works before opening the rental system with SOL payments. Build first. Rent later.