LangSmith Deployment

LangSmith Deployment is a workflow orchestration runtime purpose-built for agent workloads. It provides the managed infrastructure agents need to run reliably in production at scale, supporting the full lifecycle from local development to deployment.

Get started

Create an account

Create an API key

Go to your Settings page → API Keys → Create API Key. Copy the key and save it securely.

Choose your integration

LangSmith works with many frameworks and providers. Browse available integrations to connect your stack.

Deployable products

LangSmith Deployment is framework-agnostic which means you can deploy agents built with:

Deep Agents

Create, run, and operate Managed Deep Agents through the private preview API.

LangGraph (and LangChain)

Use the LangGraph CLI and app templates to deploy a LangGraph application to LangSmith.

Google ADK

Deploy Google Agent Development Kit (ADK) agent as a LangGraph with the deployments-wrap-sdk package.

Other frameworks

Use the LangGraph Functional API to deploy Strands, CrewAI, and other agent frameworks.

Deployment environments

You can run the same Agent Server runtime in several hosting models. A standalone server is the lightest option: you run containers yourself without the LangSmith control plane. For managed deployments through the UI and APIs, use Cloud or Self-hosted (full platform in your infrastructure).

Cloud

Fully managed by LangChain, running on AWS and GCP. Create deployments from GitHub in the LangSmith UI or with langgraph deploy. Requires a Plus plan or above.

Standalone server

Deploy Agent Server with Docker, Compose, or Kubernetes. Bring your own PostgreSQL, Redis, and LangSmith license; no control plane. Optional LangSmith tracing to Cloud or a self-hosted instance.

Self-hosted

Run the full LangSmith platform, including the control plane and data plane, in your cloud (for example on Kubernetes). Requires Enterprise plan. Integrates observability, evaluation, and agent deployment in one private stack.

Same runtime, same APIs. What changes is who manages the infrastructure. For a feature-level comparison and infrastructure setup, see Platform setup.

Deployment capabilities

Once an agent is deployed, you work with Agent Server’s execution model: assistants for configuration, threads for state, and runs for workloads.

Core capabilities

Stream to users, pause for human review, handle concurrent input, and connect via MCP and A2A.

Studio

Use an interactive environment for developing and debugging agents.

Advanced configuration

Authentication, encryption, custom routes, and short- and long-term memory stores.

Agent composition

RemoteGraph lets any agent call other deployed agents with MCP and A2A.

Find and fix failures with Engine

Once agents are in production, use LangSmith Engine to detect recurring failures in their traces, diagnose root causes, and resolve them.

Reference & operations

Tutorials

Collect user feedback for Agent Server runs: Attach end-user feedback to runs and traces
Deploy other frameworks (e.g., Strands, CrewAI): Wrap existing agents with Functional API and deploy
Implement generative user interfaces with LangGraph: Stream UI elements to a React client
Implement a CI/CD pipeline: Automate tests, evaluations, and deployments with GitHub Actions

Securing and customizing your server

Custom auth: Authentication and multi-tenant access control
Server customization: Custom routes, middleware, lifespan hooks, encryption

Operations

CI/CD pipelines
TTL configuration for state and thread management
Semantic search