Hacker News
LLMCap: Open-Source Proxy That Hard-Stops LLM API Calls at a Dollar Cap — No More Surprise Bills
LLMCap is an open-source FastAPI + Redis proxy that enforces hard dollar spending caps on LLM API calls. Unlike alerts, it actively kills the connection mid-stream when the budget is exceeded — the triggering token is never billed. The managed service at proxy.llmcap.io runs with under 35ms latency worldwide. Existing error handling works as-is since the proxy returns standard 429 responses — 21 points, 16 comments on HN.
↳ Follow the thread