⬢github Go · 7 ★ +4 since we first saw it · pushed 2 d ago · Apache-2.0
garritfra/llmtee
Self-hosted LLM proxy
llmtee is a self-hosted proxy that sits between your LLM clients and providers like OpenAI, Anthropic, OpenRouter, Ollama, or vLLM. Like Unix tee, it forwards responses unchanged while logging every call with tokens, cost, latency, and time to first token. It supports per-key daily/monthly budgets (returning 429 when exceeded), stores data in SQLite or PostgreSQL, and exposes logs via a web UI, CSV, Prometheus, or MCP.
Why now: It was just launched on Hacker News as a 'Show HN' demonstrating logging of every LLM call with tokens, cost, latency, and time to first token.
Who it is for: Individuals and small teams who want visibility into, and spending caps for, their LLM API usage without a full enterprise gateway.
llmproxy
Stars over our 22 snapshots: 3 to 7, since 5 h ago.
Where people talked about it
API: https://socialmediatrends-api.osmike.com/v1/repos/garritfra/llmtee