ai-serving-cost v0.1.0 — Serving cost math for AI workloads: per-token dollar cost with cache discounts, request cost from token counts, and what the cached share of traffic saves at current volume.

Sign in to join the discussion — your thebay.events account works here.