langfuse

Use the langfuse.openai drop-in replacement for the OpenAI client, which automatically traces all calls. You can pass extra parameters like name, session_id, user_id, tags, and metadata, and it works with streaming and async calls.

Create a Langfuse CallbackHandler with your API keys and pass it via the config callbacks parameter when invoking chains, agents, or retrievers. You can also set it as the default handler so all LangChain calls are traced automatically.

Yes, but traces are batched, so you must call langfuse.flush() before the function exits or data will be lost. For critical traces, consider synchronous mode or context managers to guarantee delivery.

Traces without user_id and session_id cannot be grouped by user or session, limiting debugging and analytics. Always pass consistent identifiers and relevant metadata when creating traces or generations.

Self-hosting requires managing your own infrastructure, and high-volume workloads may need optimization. The real-time dashboard has some latency, and evaluation features require additional setup.