Introduction to LangSmith Deployment (LangChain Academy)
by LangChain
Take a LangGraph or Deep Agents tutor agent from local code to a deployed, authenticated, memory-backed service on LangSmith.
Overview
Introduction to LangSmith Deployment is a free course from LangChain Academy, LangChain's own training site, on running agents in production with LangSmith Deployment. It has 37 lessons in a welcome section, six modules and a conclusion, with parallel setup lessons for Python and TypeScript. Module 1 builds and deploys a tutor agent. Module 2 covers connecting to the deployed agent: the client, assistants, wiring up a UI and adding custom routes. Module 3 walks through the deployment UI. Module 4 is on memory, covering memory in LangGraph, accessing the storage layer and memory in Deep Agents. Module 5 covers multi-tenant auth: authentication, authorization, plugging in an auth provider and adding auth to the tutor. Module 6 deploys a Deep Agent without writing deployment code, using the deepagents-cli, with an exercise to deploy the tutor and a lesson on bringing your own UI. The companion repository, langchain-ai/lca-deployment on GitHub (MIT licence, 132 commits), has Python and TypeScript tracks. The Python track needs Python 3.11 to 3.13, uv, a LangSmith API key and an Anthropic API key, and Docker is optional for the `langgraph deploy` build. Local deployment works on a free LangSmith Developer account. Cloud deployment requires a Plus plan or above.
At a Glance
- Topic
- Agentic
- Level
- Intermediate
- Format
- Course
- Cost
- Free
- Duration
- 37 lessons across 6 modules, self-paced; no official length given, and the page lists 0 hours of video, so expect a text-and-code course
- Provider
- LangChain
- Hands-on
- Yes — code/exercises
- Certificate
- None
What You’ll Learn
- ✓Deploy a LangGraph tutor agent to LangSmith Deployment and call it
- ✓Connect clients to a deployed agent through the SDK and assistants
- ✓Wire a chat UI and custom HTTP routes onto a deployed agent server
- ✓Use the LangSmith deployment UI to inspect and manage running agents
- ✓Add long-term memory with the LangGraph store and Deep Agents memory
- ✓Implement multi-tenant authentication and authorization with an external auth provider
- ✓Deploy a Deep Agent from the deepagents-cli without writing deployment code
- ✓Choose between local deployment on a free account and cloud deployment
Highlights
- •Covers the parts tutorials skip after the agent works: auth, per-user memory, custom routes and a UI
- •Every module has both Python and TypeScript setup, backed by the official lca-deployment repo
- •One tutor agent runs through all six modules, so each lesson extends working code
- •Free to take, and local deployment runs on a free LangSmith Developer account
- •Ends on the deepagents-cli path, which deploys a Deep Agent with no deployment code
Who It’s For
Best For
- ✓Developers with a working LangGraph or Deep Agents prototype who need to put it behind an API
- ✓Platform engineers evaluating LangSmith Deployment as an agent hosting layer
- ✓Teams building multi-user agents that need per-tenant auth and memory isolation
Prerequisites
- •Working knowledge of LangGraph or LangChain, for example from the Academy's Introduction to LangGraph
- •Python 3.11 to 3.13 with uv, or the TypeScript track's toolchain
- •A LangSmith API key and an Anthropic API key; a LangSmith Plus plan for cloud deployment
- •Docker is optional but speeds up the langgraph deploy build
FAQ
What is Introduction to LangSmith Deployment (LangChain Academy)?
Introduction to LangSmith Deployment is a free LangChain Academy course for developers who already have a LangGraph or Deep Agents agent working locally and need to run it as a service. You deploy a tutor agent, connect clients and a UI to it, add long-term memory and multi-tenant auth, then ship a Deep Agent with the deepagents-cli.
Is Introduction to LangSmith Deployment (LangChain Academy) free?
Introduction to LangSmith Deployment (LangChain Academy) is free to access.
What level is Introduction to LangSmith Deployment (LangChain Academy) for?
Introduction to LangSmith Deployment (LangChain Academy) is aimed at a intermediate audience. Recommended background: Working knowledge of LangGraph or LangChain, for example from the Academy's Introduction to LangGraph, Python 3.11 to 3.13 with uv, or the TypeScript track's toolchain, A LangSmith API key and an Anthropic API key; a LangSmith Plus plan for cloud deployment, Docker is optional but speeds up the langgraph deploy build.
How long does Introduction to LangSmith Deployment (LangChain Academy) take?
Expect roughly 37 lessons across 6 modules, self-paced; no official length given, and the page lists 0 hours of video, so expect a text-and-code course. Most learners work through it at their own pace.
What will I learn from Introduction to LangSmith Deployment (LangChain Academy)?
You'll learn: Deploy a LangGraph tutor agent to LangSmith Deployment and call it; Connect clients to a deployed agent through the SDK and assistants; Wire a chat UI and custom HTTP routes onto a deployed agent server; Use the LangSmith deployment UI to inspect and manage running agents; Add long-term memory with the LangGraph store and Deep Agents memory; Implement multi-tenant authentication and authorization with an external auth provider; Deploy a Deep Agent from the deepagents-cli without writing deployment code; Choose between local deployment on a free account and cloud deployment.
Topics
Sources
This page was written from 4 sources, 2 on domains other than academy.langchain.com.