Skip to main content

Using Gemini API in LangChain: Step-by-Step Tutorial

What is LangChain and Why Use It?

LangChain is an open-source framework that simplifies the use of Large Language Models (LLMs) like OpenAI, Gemini (Google), and others by adding structure, tools, and memory to help build real-world applications such as chatbots, assistants, agents, or AI-enhanced software.

Why Use LangChain for LLM Projects?

  • Chainable Components: Easily build pipelines combining prompts, LLMs, tools, and memory.
  • Multi-Model Support: Work with Gemini, OpenAI, Anthropic, Hugging Face, etc.
  • Built-in Templates: Manage prompts more effectively.
  • Supports Multi-Turn Chat: Manage complex interactions with memory and roles.
  • Tool and API Integration: Let the model interact with external APIs or functions.

Let's Walk Through the Code: Gemini + LangChain

I will break the code into 4 main parts, each showcasing different features of LangChain and Gemini API.

Part 1: Basic Gemini API Call Using LangChain

import os
from dotenv import load_dotenv
load_dotenv(".env")  # Load environment variables from .env file
api_key = os.getenv("GOOGLE_API_KEY")  # Get Google API Key securely

# Import Gemini LLM wrapper from LangChain
from langchain_google_genai import ChatGoogleGenerativeAI

# Initialize the Gemini model with specific configuration
llm = ChatGoogleGenerativeAI(model='gemini-2.0-flash', temperature=0.9)

# Send a prompt to the model and get a response
response = llm.invoke('Explain what ChatGoogleGenerativeAI function do in langchain_google_genai.')
print(response.content)

Function Explanation:

  • load_dotenv(".env"): Loads your .env file which stores your API key.
  • os.getenv(): Retrieves your Gemini API key without hardcoding it in the script.
  • ChatGoogleGenerativeAI(): Initializes the Gemini model with the name and temperature (creativity level).
  • invoke(): Sends a prompt to the LLM and returns a response object.

This is a basic setup that demonstrates how to call Gemini via LangChain and print the result.

Part 2: Using PromptTemplate and Chain Composition

from langchain.prompts import PromptTemplate

llm = ChatGoogleGenerativeAI(model='gemini-2.0-flash', temperature=0.9)

# Define a reusable prompt structure with a placeholder
prompt = PromptTemplate.from_template('You are python coding agent. Explain me how to use {message}.')

# Chain the prompt template with the model using the pipe operator
chain = prompt | llm

# Provide the actual input for the placeholder
message = 'heapq'
response = chain.invoke(input=message)
print(response.content)

Function Explanation:

  • PromptTemplate.from_template(): Creates a parameterized prompt where {message} is replaced dynamically.
  • | (pipe operator): Chains the prompt and the LLM, forming a pipeline: prompt → LLM.
  • invoke(input=message): Passes the actual user input ('heapq') to the chain and gets the final output.

This part shows how to modularize your prompts and chain components together, which is a powerful pattern in LangChain.

Part 3: Using System and Human Messages for Role-Based Input

from langchain_core.messages import HumanMessage, SystemMessage

llm_pro = ChatGoogleGenerativeAI(model='gemini-2.5-pro-exp-03-25', temperature=0.9)

# Create structured messages indicating who is speaking (system vs human)
output = llm_pro.invoke([
    SystemMessage(content='Answer within 100 charicter in Korean.'),
    HumanMessage(content='Forcasting Amazon stock price for next week.')
])
print(output.content)

Function Explanation:

  • SystemMessage(): Gives the model background behavior (e.g., how to answer).
  • HumanMessage(): Simulates the user’s actual question or prompt.
  • Passing a list of messages: Supports multi-turn conversation or giving context with different roles.

This part introduces the chat-based message structure—important for building assistants with clear roles and behavior instructions.

Part 4: Using ChatPromptTemplate with Role Keywords

from langchain_core.prompts import ChatPromptTemplate

llm_pro = ChatGoogleGenerativeAI(model='gemini-2.5-pro-exp-03-25', temperature=0.9)

# Define a chat-style prompt template with role-based content
prompt = ChatPromptTemplate([
    ('system', 'Answer within 100 charicter in Korean.'),
    ('user', 'Forcasting {message} stock price for next week.')
])

# Create a chain from the prompt to the model
chain = prompt | llm_pro

# Replace the variable in the prompt template
message = 'Amazon'
response = chain.invoke(input=message)
print(response.content)

Function Explanation:

  • ChatPromptTemplate(): Provides a flexible way to define role-specific messages using keywords like 'system' and 'user'.
  • Templates with variables: {message} is dynamically filled in at runtime.
  • Prompt Chaining: Works just like before—template flows into the LLM.

This version is cleaner and more scalable than manually defining SystemMessage and HumanMessage.

2025.04.23 - [AI] - Building a Simple LLM Chatbot Server Using Google Gemini 2.5 Pro and LangChain

References

Tags

Comments

Post a Comment

Popular

Understanding SentencePiece: A Language-Independent Tokenizer for AI Engineers

In the realm of Natural Language Processing (NLP), tokenization plays a pivotal role in preparing text data for machine learning models. Traditional tokenization methods often rely on language-specific rules and pre-tokenized inputs, which can be limiting when dealing with diverse languages and scripts. Enter SentencePiece—a language-independent tokenizer and detokenizer designed to address these challenges and streamline the preprocessing pipeline for neural text processing systems. What is SentencePiece? SentencePiece is an open-source tokenizer and detokenizer developed by Google, tailored for neural-based text processing tasks such as Neural Machine Translation (NMT). Unlike conventional tokenizers that depend on whitespace and language-specific rules, SentencePiece treats the input text as a raw byte sequence, enabling it to process languages without explicit word boundaries, such as Japanese, Chinese, and Korean. This approach allows SentencePiece to train subword models di...
Post a Comment
Read more

Mastering the Byte Pair Encoding (BPE) Tokenizer for NLP and LLMs

Byte Pair Encoding (BPE) is one of the most important and widely adopted subword tokenization algorithms in modern Natural Language Processing (NLP), especially in training Large Language Models (LLMs) like GPT. This guide provides a deep technical dive into how BPE works, compares it with other tokenizers like WordPiece and SentencePiece, and explains its practical implementation with Python code. This article is optimized for AI engineers building real-world models and systems. 1. What is Byte Pair Encoding? BPE was originally introduced as a data compression algorithm by Gage in 1994. It replaces the most frequent pair of bytes in a sequence with a single, unused byte. In 2015, Sennrich et al. adapted BPE for NLP to address the out-of-vocabulary (OOV) problem in neural machine translation. Instead of working with full words, BPE decomposes them into subword units that can be recombined to represent rare or unseen words. 2. Why Tokenization Matters in LLMs Tokenization is th...
Post a Comment
Read more

ZeRO: Deep Memory Optimization for Training Trillion-Parameter Models

In 2020, Microsoft researchers introduced ZeRO (Zero Redundancy Optimizer) via their paper "ZeRO: Memory Optimization Towards Training Trillion Parameter Models" (arXiv:1910.02054). ZeRO is a memory optimization technique that eliminates redundancy in distributed training, enabling efficient scaling to trillion-parameter models. This provides an in-depth technical breakdown of ZeRO's partitioning strategies, memory usage analysis, and integration with DeepSpeed. 1. What is ZeRO? ZeRO eliminates redundant memory copies of model states across GPUs. Instead of replicating parameters, gradients, and optimizer states across each GPU, ZeRO partitions them across all devices. This results in near-linear memory savings as the number of GPUs increases. 2. Limitations of Traditional Data Parallelism In standard data-parallel training, every GPU maintains: Model Parameters $\theta$ Gradients $\nabla \theta$ Optimizer States $O(\theta)$ This causes memory usage ...
Post a Comment
Read more