LangSmith
Free tierThe end-to-end platform for building, testing, and monitoring LLM-powered applications
Free tier available·Technical·Powered by LangChain (model-agnostic)·API available
Key strengths
Deep LangChain & LangGraph integration for zero-config tracingEnd-to-end LLM observability with detailed trace visualizationBuilt-in evaluation and dataset management for regression testingPrompt management with versioning and A/B testing supportProduction monitoring with feedback collection and alerting
Free tier + paid plans
San Francisco, USA
Founded 2023
Self-hostable
No ratings yet
Developer Integration Guide
Installation
pip install langsmith # Python
npm install langsmith # TypeScript/Node
Tracing
LangSmith auto-instruments LangChain and LangGraph applications via environment variables. For non-LangChain code, use the @traceable decorator (Python) or traceable() wrapper (TS):
from langsmith import traceable
@traceable
def my_llm_call(prompt: str) -> str:
# your OpenAI / Anthropic / etc. call here
return response
Evaluations & Datasets
Use the Client class to create datasets and run evaluators:
from langsmith import Client
from langsmith.evaluation import evaluate
client = Client()
dataset = client.create_dataset("my-dataset")
client.create_examples(inputs=[{"q": "..."}], outputs=[{"a": "..."}], dataset_id=dataset.id)
results = evaluate(
my_pipeline,
data="my-dataset",
evaluators=[correctness_evaluator],
)
CI/CD Integration
Trigger evaluation suites programmatically or via the LangSmith GitHub Action to gate deployments on LLM quality metrics.
Self-Hosting
LangSmith can be self-hosted via Helm charts on Kubernetes or Docker Compose for air-gapped / enterprise environments. See the self-hosting docs for full configuration.