LangSmith has introduced full end-to-end OpenTelemetry support for applications built on LangChain and LangGraph, enhancing observability for large language model (LLM) applications. This update integrates OpenTelemetry into the LangSmith SDK, enabling standardized tracing across different platforms, including LangSmith's own testing and observability platform, as well as other systems like Datadog and Grafana. OpenTelemetry, an open-source framework, standardizes the collection, export, and analysis of telemetry data, addressing unique challenges in LLM observability by providing a consistent way to monitor complex workflows and dynamic outputs. LangSmith's OpenTelemetry pipeline allows users to automatically generate detailed traces from LangChain or LangGraph applications, convert and transport these traces through the LangSmith SDK, and visualize them in a specialized dashboard. While the integration offers interoperability and a unified view of application stacks, users focused solely on LangSmith might prefer its native tracing format for optimal performance due to its lower computational and memory overhead tailored for LLM applications.