Skip to main content
Before diving in, note that LangChain evolves quickly alongside generative AI. This guide covers versions 0.1.10 and 0.1.11. To match the examples below, install one of these versions: 
pip install langchain==0.1.11
# or
pip install langchain==0.1.10
Always pin your LangChain version to ensure compatibility with course notebooks and examples.

​
Official Website and Documentation

Start your LangChain journey by browsing the official site and documentation:
The image shows two screenshots of the LangChain website and documentation, featuring sections on OpenAI integration and an introduction to LangChain.
You can also follow the LangChain blog for tutorials and release notes, and explore the Oracle Cloud Infrastructure Python SDK docs:
The image shows two website screenshots: one of the Oracle Cloud Infrastructure Documentation for Python SDK and the other of the LangChain blog page.

​
Documentation Overview

LangChain’s documentation covers everything from basic concepts to advanced modules:
The image shows a webpage from LangChain's documentation, featuring a navigation menu on the left and a diagram explaining LangChain's components and libraries in the center.
Below is a quick reference for core modules aligned with this course:
ModuleDescription
Model I/OFormatting, predicting, and parsing LLM requests
Prompt EngineeringBuilding and testing templates
Chat ModelsConversational interfaces
Output ParsersStructured data extraction
Retrieval AgentsQuerying external knowledge
ChainsOrchestrating multi-step processes
MemoryContext management between interactions

​
Core Modules and Model I/O

LangChain’s core sections include Model I/O, prompt engineering, chat models, output parsers, retrieval agents, chains, and memory. Here’s a representative flowchart for Model I/O:
The image shows a webpage from LangChain's documentation, specifically the "Model I/O" section, featuring a flowchart illustrating the process of formatting, predicting, and parsing data with language models.
New modules such as LangServ, LangSmith, and LangGraph are under active development and not covered in this guide. Apply for early access if you’d like to experiment.

​
Third-Party Integrations

LangChain integrates with dozens of LLM providers, embedding models, and vector stores. You can filter integrations based on support for invoke, async, streaming, batch, and more:
The image shows a webpage from LangChain's documentation, detailing LLM integration features with a table indicating support for various functionalities like invoke, async invoke, stream, and batch for different models.
Embedding models and databases follow a similar pattern—check the Integrations section for your preferred vendor.

​
Python SDK and API Reference

Every LangChain component is documented under the API reference. You’ll find details for agents, language models, chains, toolkits, and community modules.

​
Agents

The image shows a webpage from the LangChain documentation, specifically focusing on the "langchain.agents" section, detailing classes and functions related to agents in version 0.1.11.

​
Language Models

The image shows a webpage from the LangChain documentation, specifically focusing on the "langchain_core.language_models" module, detailing class hierarchies, main helpers, classes, and functions related to language models.

​
Agent Toolkits

A variety of toolkits help you build and customize agents:
The image shows a webpage from the LangChain documentation, listing various toolkits and functions related to agent toolkits, along with brief descriptions of each.

​
Community LLMs and Modules

Community contributions extend core functionality. Browse community LLM implementations and modules:
The image shows a webpage from the LangChain documentation, specifically the section on community LLMs (Large Language Models), listing various classes and their descriptions. The left sidebar contains a navigation menu with different modules.
The image shows a webpage with a list of LangChain community modules and their descriptions, focusing on various large language models and integrations.

​
Code Examples

​
Initializing an OpenAI Chat Model

from langchain_community.llms import OpenAIChat

openai_chat = OpenAIChat(model_name="gpt-3.5-turbo")

​
Creating an LLMChain

from langchain.chains import LLMChain
from langchain_community.llms import OpenAI
from langchain_core.prompts import PromptTemplate

prompt = PromptTemplate(
    input_variables=["adjective"],
    template="Tell me a {adjective} joke"
)

llm_chain = LLMChain(llm=OpenAI(), prompt=prompt)

​
Exploring Chains

Discover all chain implementations in the API reference:
The image shows a webpage from the LangChain documentation, specifically the API reference for chains, listing various classes and their descriptions.

​
Blog and Updates

Stay up to date with the latest tutorials, release notes, and community announcements:

Watch Video

Practice Lab