What is Langbase Agent?
Agent, an AI Primitive by Langbase, works as a runtime LLM agent. You can specify all parameters at runtime and get the response from the agent.
Agent uses our unified LLM API to provide a consistent interface for interacting with 100+ LLMs across all the top LLM providers. See the list of supported models and providers here.
All cutting-edge LLM features are supported, including streaming, JSON mode, tool calling, structured outputs, vision, and more. It is designed to be used in a variety of applications, including agentic workflows, chatbots, virtual assistants, and other AI-powered applications.
Quickstart: Create a Runtime AI Agent
Let's get started
In this guide, we'll use the Langbase SDK to create an AI agent that can summarize user support queries.
Step #1Generate Langbase API key
Every request you send to Langbase needs an API key. This guide assumes you already have one. If not, please check the instructions below.
Step #2Setup your project
Create a new directory for your project and navigate to it.
Project setup
mkdir agent && cd agent
Initialize the project
Create a new Node.js project.
Initialize project
npm init -y
Install dependencies
You will use the Langbase SDK to run the agent and dotenv
to manage environment variables.
Install dependencies
npm i langbase dotenv
Create an env file
Create a .env
file in the root of your project. You will need two environment variables:
LANGBASE_API_KEY
: Your Langbase API key.LLM_API_KEY
: Your LLM provider API key.
.env
LANGBASE_API_KEY=your_api_key_here
LLM_API_KEY=your_llm_api_key_here
Step #2Configure and Run the agent
Now let's create a new file called agent.ts
in the root of your project. This file will contain the code and configuration of the agent.
We will use OpenAI GPT-4.1 model, but you can use any other supported model listed here.
In instructions, which are like system prompts, we will specify that the agent is a support agent and should summarize user support queries. Finally, we will provide the user query as input to the agent.
agent.ts
import { Langbase } from 'langbase';
import dotenv from 'dotenv';
dotenv.config();
// Initialize the Langbase client
const langbase = new Langbase({
apiKey: process.env.LANGBASE_API_KEY!
});
async function main() {
const response = await langbase.agent.run({
model: 'openai:gpt-4.1',
stream: false,
apiKey: process.env.LLM_API_KEY!,
instructions: 'You are an AI agent that summarizes user support queries for a support agent.',
input: 'I am having trouble logging into my account. I keep getting an error message that says "Invalid credentials." I have tried resetting my password, but it still does not work. Can you help me?',
});
console.log('Agent Response:', response.output);
}
main();
Run the agent by executing the script agent.ts
.
Run the script
npx tsx agent.ts
You should see an output similar to:
Agent Response: User can't log in. Gets "Invalid credentials" error even after password reset. Needs help.
Next Steps
Now that you have a basic understanding of how to create and run an agent, you can explore more advanced features and configurations. Here are some suggestions:
- Set
stream: true
for streaming responses (great for chat related applications) - Use structured outputs to get structured data from the agent
- Explore more code examples of agents here
- Create complex AI workflows with multiple agents in Langbase Workflow