Skip to content

emo-eth/local-model-router

v0.1.0

Tailnet-wide OpenAI-compatible local model inference router with exact model contracts, invisible placement, fail-fast capacity, fair responsiveness, and Pi-grain operator telemetry.

GitHub
local-model-router

Operate and inspect the local model inference router across home fleet devices (DGX Sparks, Windows 5090/4090, local boxes). Use when configuring local model providers, diagnosing fleet bottlenecks, checking Pi-grain inference latency, or verifying fail-fast routing.