LangChain Environment Setup

This guide will walk you through installing and configuring LangChain so you can successfully run your first LangChain program.

Environment Requirements

ProjectMinimum VersionRecommended Version
Python3.103.11 or 3.12
pip22.0Latest version
Operating SystemmacOS / Linux / Windows all supported

Check the Python version:

$ python --version
Python 3.12.7

If you haven't installed Python yet, refer to ourPython3 Environment Setup。

Reference for Python package and environment management tools:uv Getting Started Tutorial。

Installing LangChain

LangChain uses a modular design, with core functionality and third-party integrations installed separately.

Install the Core Package

$ pip install langchain

This command installs thelangchainmain package, which includes core APIs such as init_chat_model() and create_agent(), and automatically installs langchain-core as a dependency.

Install the Model Provider Package

LangChain itself does not include specific model implementations. You need to install the corresponding provider package based on the model you are using:

Model ProviderInstallation CommandModels Used
OpenAIpip install langchain-openaiGPT-4, GPT-5, etc.
Anthropicpip install langchain-anthropicClaude series
DeepSeekpip install langchain-deepseekDeepSeek-V3, R1, etc.
Googlepip install langchain-google-genaiGemini series
Ollama (local models)pip install langchain-ollamaLocal models such as Llama, Qwen, etc.
xAIpip install langchain-xaiGrok series
Mistralpip install langchain-mistralaiMistral series

One-Click Installation

Beginners are advised to directly install the following combination:

$ pip install langchain langchain-openai python-dotenv

python-dotenv is used to load API keys from the .env file. This is the best practice for managing sensitive information and avoids hardcoding the key in your code.

Configure API Key

Get an API Key

Taking OpenAI as an example, you need to register an account and obtain an API key first:

  1. Visithttps://platform.openai.comRegister or log in
  2. Go to the API Keys page and click "Create new secret key"
  3. Copy the generated key (format: sk-xxxxxxxx)

The API key is only shown once, so save it immediately in a safe place. Do not commit the key to a Git repository or share it with others.

The process for other providers is similar; please refer to their respective documentation.

Set Environment Variables

It is recommended to use a .env file to manage the API key. Create a .env file in the project root directory:

Example

# File path: .env
# Fill in your API key
OPENAI_API_KEY=sk-your-api-key-here

# If you use other models, also configure them here
# ANTHROPIC_API_KEY=sk-ant-your-key
# DEEPSEEK_API_KEY=sk-your-key

Then create a .gitignore file to ensure the .env file is not committed:

# .gitignore
.env
__pycache__/
*.pyc

Load in Code

Example

# File path: config.py
# Load the .env file at the beginning of the program
import os
from dotenv import load_dotenv

# Load environment variables from the .env file
load_dotenv()

# Verify whether the API key was loaded successfully
api_key = os.getenv("OPENAI_API_KEY")
if api_key:
    # Only display the first 8 and last 4 characters to avoid leaking the full key
    print(f"API key loaded: {api_key[:8]}...{api_key[-4:]}")
else:
    print("Warning: OPENAI_API_KEY not found. Please check the .env file")

Verify Installation

Run the following script to verify whether the installation was successful:

Example

# File path: verify_install.py
# Verify LangChain installation and API key configuration
from dotenv import load_dotenv
load_dotenv()

# Test 1: Verify langchain import
try:
    import langchain
    print(f"langchain version: {langchain.__version__}")
except ImportError:
    print("Error: langchain is not installed. Please run pip install langchain")

# Test 2: Verify langchain-openai import
try:
    import langchain_openai
    print("langchain-openai is installed")
except ImportError:
    print("Error: langchain-openai is not installed. Please run pip install langchain-openai")

# Test 3: Verify API key configuration
import os
api_key = os.getenv("OPENAI_API_KEY")
if not api_key:
    print("Error: OPENAI_API_KEY is not configured. Please set it in the .env file")
else:
    print(f"API key configured successfully: {api_key[:8]}...{api_key[-4:]}")

# Test 4: Send a test request
from langchain.chat_models import init_chat_model

model = init_chat_model("openai:gpt-4o-mini")
response = model.invoke("Introduce Python in one sentence.")
print(f"\nModel reply: {response.content}")

Output:

langchain 版本: 1.3.0
langchain-openai 已安装
API Key 配置成功: sk-proj-z...xxxx
Model reply: Example is a Chinese technical learning platform for programming beginners.

If the last step reports an error, check: (1) whether the API key is correct; (2) whether a network proxy is affecting the connection; (3) whether the account balance is sufficient.

In China, we can use the DeepSeek large model for testing. If you don't have one yet, you need to firsthttps://platform.deepseek.com/api_keyscreate an API key.

DeepSeek's API documentation reference:https://api-docs.deepseek.com/zh-cn/。

If you need to manage multiple third-party models in a unified way, you can choose to install LiteLLM as a model gateway:

pip install -U litellm

However, the examples in this article directly useChatOpenAItogether withopenai_api_baseparameters to connect to DeepSeek, so there is no need to install LiteLLM additionally.

Example

import os
from langchain_openai import ChatOpenAI

# It is recommended to put the API key in an environment variable, or assign it directly here (do not hardcode it in production)
# os.environ["DEEPSEEK_API_KEY"] = "sk-your-DeepSeek-key"

# Initialize the model
llm = ChatOpenAI(
    model="deepseek-v4-pro",             # DeepSeek V4 model name
    openai_api_key="sk-your-Key",        # Fill in your DeepSeek API key
    openai_api_base="https://api.deepseek.com", # DeepSeek API endpoint
    temperature=0.7,
    max_tokens=1024
)

# Let's test it
response = llm.invoke("Hello, DeepSeek! Please give a brief self-introduction.")
print(response.content)

After executing the above code, the output will be:

你好！很高兴认识你！

我是DeepSeek，由深度求索公司创造的AI助手。我的特点包括：
。。。

Suggested Project Directory

Recommended project directory structure:

langchain-learning/
├── .env                  # API Key 配置（不提交到 Git）
├── .gitignore            # Git 忽略规则
├── config.py             # 公共配置加载
├── 01_hello_world.py     # 第一篇的示例
├── 02_first_agent.py     # Agent 示例
└── ...                   # 更多练习
Other Extensions