LLMCap

LLMCap is a reverse proxy designed to enforce strict spending limits on large language model API calls. Its core features include: * Hard dollar caps for API usage * Support for major providers: Anthropic, OpenAI, Gemini, Mistral, Cohere * Minimal added latency to API requests * Live spend dashboard and comprehensive audit logs * Simple integration with a single base_url change This tool ensures that once your predefined spending limit is reached, all subsequent API requests are immediately blocked with a 429 status code, preventing any further charges from the provider. This proactive capping mechanism eliminates the risk of unexpected or runaway expenses, providing financial control for your projects. Your API keys remain secure, passing through per-request without being stored. LLMCap provides flexible capping options, allowing users to define daily, monthly, or per-key dollar limits with granularity down to individual models. It integrates seamlessly into various development workflows, offering a VS Code Extension for live spend monitoring, a Terminal CLI for command-line management, and a Windows Tray App for constant oversight. It's an essential utility for developers, teams, and enterprises seeking to manage and control their expenditures effectively when interacting with external language model services.