A reverse proxy that enforces hard dollar caps on LLM API calls. Change one line (base_url), set a cap — when you hit it, the next request returns 429 before the token reaches the provider. Money never spent. → Anthropic, OpenAI, Gemini, Mistral, Cohere → <35ms added latency → Live spend dashboard, audit logs Starter $19/mo · Pro $49/mo · 3-day trial.
Hey PH! I built LLMCap after seeing a $37,901 AWS Bedrock bill from a single misconfiguration posted on GitHub.
The problem with existing tools (Helicone, AWS Budgets) is they alert you AFTER the money is spent. LLMCap is a reverse proxy — when you hit your cap, the request returns 429 before the token ever reaches the provider.
The interesting engineering challenge: enforcing caps mid-stream without buffering the entire response. Happy to answer questions about the architecture!
Report
Finally something that actually stops the request instead of just alerting you after. This is exactly what agent developers need.
Report
No reviews yetBe the first to leave a review for LLMCap
Finally something that actually stops the request instead of just alerting you after. This is exactly what agent developers need.