Langfuse
Open-source LLM engineering platform. Tracing, evals, prompt management and cost tracking for LLM apps.
- #llm
- #observability
- #tracing
- #evals
- #open-source
Langfuse records traces of your LLM application: every model call, retrieval step and tool call, with latency, token usage and cost attached.
Why it’s useful
- Debug RAG: see exactly which chunks were retrieved for a query and what the model did with them.
- Track spend: actual token usage per trace. Compare it against estimates from the RAG Chunk & Token Calculator.
- Prompt management and evals built in. Open source (MIT) and self-hostable.