Inferock Bench is a local diagnostic proxy that sits between your applications and LLM providers (OpenAI, Anthropic, Gemini, and pinned OpenRouter endpoints). It captures detailed per-call event records, including token usage, pricing evidence, delivery failures, and retry signals. Using the Inferock Standard, it generates independent receipts that separate observed spend from computed money loss, time loss, and invoice-check exposure. This gives you a transparent, auditable view of what you're actually billed for, helping you spot failures, overcharges, and billing discrepancies. It runs entirely on your machine, keeping your provider keys local, and is designed for developers and teams who want to verify their AI costs.