🔭 LangSmith

2 guides covering common problems, patterns, and production issues in LangSmith.

LangSmith is LangChain's production observability and evaluation platform for LLM applications. It provides distributed tracing via the @traceable decorator, a prompt Hub for version-controlled prompts, LLM-as-judge evaluations, dataset experiments, and cost and latency dashboards.

  • Distributed tracing with @traceable and LangChain callbacks
  • Prompt Hub — version control, A/B test, and deploy prompts
  • LLM-as-judge automated quality evaluations
  • Dataset experiments for regression testing and evals
  • Cost and latency dashboards by model, user, and tag
Visit official site →

Stay sharp as AI tools evolve

New guides drop regularly. Get them in your inbox — no noise, just signal.