Introducing End-to-End OpenTelemetry Support in LangSmith

LangSmith now supports full end-to-end OpenTelemetry (OTel) for LangChain and LangGraph applications, ensuring seamless observability from your application code to the LangSmith platform.

Why It Matters

With our OTel support, you'll be able to get::

• Unified observability: See your entire application stack in one place

• Distributed tracing: Track requests across microservices

• Interoperability: Integrate with existing observability tools via OpenTelemetry standards

Getting Started

1. Install OpenTelemetry support:

   pip install "langsmith[otel]" langchain

2. Enable OpenTelemetry by setting environment variables:

   LANGSMITH_OTEL_ENABLED=true
   LANGSMITH_TRACING=true
   LANGSMITH_ENDPOINT=https://api.smith.langchain.com
   LANGSMITH_API_KEY=<your_langsmith_api_key>

3. Use tracing in your LangChain app:

   from langchain_openai import ChatOpenAI
   from langchain_core.prompts import ChatPromptTemplate
   
   prompt = ChatPromptTemplate.from_template("Tell me a joke about {topic}")
   model = ChatOpenAI()
   chain = prompt | model
   result = chain.invoke({"topic": "programming"})
   print(result.content)

Try It Today

Get started with OpenTelemetry in LangSmith— check out our docs for more details: https://docs.smith.langchain.com/observability/how_to_guides/trace_langchain_with_otel