LangSmith is a unified observability & evals platform where teams can debug, test, and monitor AI app performance — whether building with LangChain or not.
.svg)
Quickly debug and understand non-deterministic LLM app behavior with tracing. See what your agent is doing step by step —then fix issues to improve latency and response quality.
Evaluate your app by saving production traces to datasets — then score performance with LLM-as-Judge evaluators. Gather human feedback from subject-matter experts to assess response relevance, correctness, harmfulness, and other criteria.
Experiment with models and prompts in the Playground, and compare outputs across different prompt versions. Any teammate can use the Prompt Canvas UI to directly recommend and improve prompts.
Track business-critical metrics like costs, latency, and response quality with live dashboards — then drill into the root cause when problems arise.
Yes! Many companies who don’t build with LangChain/LangGraph use LangSmith. You can log traces to LangSmith via the Python SDK, the TypeScript SDK, or the API. See here for more information.
Getting started on LangSmith requires just two environment variables in your LangChain or LangGraph code. See how to send traces from your LangGraph agent or your LangChain app.
Yes, you can log traces to LangSmith using a standard OpenTelemetry client to access all LangSmith features, including tracing, running evals, and prompt engineering. See the docs.
LangSmith traces contain the full information of all the inputs and outputs of each step of the application, giving users full visibility into their agent or LLM app behavior. LangSmith also allows users to instantly run evals to assess agent or LLM app performance — including LLM-as-Judge evaluators for auto-scoring and the ability to attach human feedback. Learn more.
Yes, we allow customers to self-host LangSmith on our enterprise plan. We deliver the software to run on your Kubernetes cluster, and data will not leave your environment. For more information, check out our documentation.
For Cloud SaaS, traces are stored in GCP us-central-1 or GCP europe-west4, depending on your plan. Learn more.
No, LangSmith does not add any latency to your application. In the LangSmith SDK, there’s a callback handler that sends traces to a LangSmith trace collector which runs as an async, distributed process. Additionally, if LangSmith experiences an incident, your application performance will not be disrupted.
We will not train on your data, and you own all rights to your data. See LangSmith Terms of Service for more information.
See our pricing page for more information, and find a plan that works for you.
Get started with LangChain, LangSmith, and LangGraph to enhance your LLM app development, from prototype to production.