Skip to main content
In this lesson, we’ll dive into prompt templates—parameterized messages that make your LangChain interactions more flexible, consistent, and reusable. Instead of hard-coding each system, human, or AI message, you define templates with placeholders that are filled in at runtime.

​
What Are Prompt Templates?

A prompt template is a blueprint for your LLM interactions. You can:
  • Define system, human, or AI messages with named placeholders
  • Populate placeholders at runtime, generating dynamic prompts
  • Chain multiple templates together, passing data from one step to the next
  • Standardize messaging across your team for a consistent LLM experience
Always verify that all placeholders match the input keys you supply at runtime to avoid template rendering errors.

​
Benefits of Using Prompt Templates

BenefitDescriptionExample
ReusabilityCreate once, reuse across multiple chainsgreeting = "Hello, {user_name}!"
ConsistencyEnforce a standard messaging formatSystem prompts that always start with company branding
FlexibilityQuickly swap out placeholders or default valuesChanging {user_name} to {customer_name} globally
MaintainabilityEasier updates when requirements evolveUpdate tone or style in one template file

​
Core Components

  1. System Message Template
    Typically sets the overall behavior of the LLM.     
    from langchain.prompts import SystemMessagePromptTemplate

system_template = SystemMessagePromptTemplate.from_template(
    "You are a helpful assistant for {company_name} support."
)
  2. Human Message Template
    Captures user input or queries.     
    from langchain.prompts import HumanMessagePromptTemplate

human_template = HumanMessagePromptTemplate.from_template(
    "Hi, my order #{order_id} is delayed. Can you help?"
)
  3. AI Message Template
    Formats or parses the LLM’s response.     
    from langchain.prompts import AIMessagePromptTemplate

ai_template = AIMessagePromptTemplate.from_template(
    "The estimated delivery date for order {order_id} is {delivery_date}."
)
The image is a diagram illustrating the flow of an AI message between an application and a chat model, using a prompt template.

​
Demo Scenarios

In the following demos, we’ll cover:
  1. Creating prompt templates with placeholders
  2. Generating parameterized prompts
  3. Building reusable, organization-wide templates
Let’s jump into the hands-on examples and see these prompt templates in action!

Watch Video