An open-source LLM observability proxy: add one line pointing your existing API calls through it and it logs every prompt and completion, with dashboards for cost, latency and usage per user or feature.
Pricing: Open-source