langbase

Langbase SDK

The AI framework for building declarative and composable AI-powered LLM products.

Getting Started with langbase SDK

Installation

First, install the langbase package using npm or yarn:

npm install langbase

or

pnpm add langbase

or

yarn add langbase

Usage

You can generateText or streamText based on the type of a pipe.

Check our SDK documentation for more details.

Example projects

Check the following examples:

• Node: Generate Text
• Node: Stream Text
• Next.js Example
  • TypeScript code
  • React component to display the response
  • API Route handlers to send requests to ⌘ Langbase

Node.js Example Code

Node.js Examples

Add a .env file with your Pipe API key

# Add your Pipe API key here.
LANGBASE_PIPE_API_KEY="pipe_12345`"

---

Generate text generateText()

For more check the API reference of generateText()

import 'dotenv/config';
import {Pipe} from 'langbase';

// 1. Initiate the Pipe.
const pipe = new Pipe({
	// Make sure you have a .env file with any pipe you wanna use.
	// As a demo we're using a pipe that has less wordy responses.
	apiKey: process.env.LANGBASE_PIPE_API_KEY!,
});

// 3. Generate the text by asking a question.
const result = await pipe.generateText({
	messages: [{role: 'user', content: 'Who is an AI Engineer?'}],
});

// 4. Done: You got the generated completion.
console.log(result.completion);

---

Stream text streamText()

For more check the API reference of streamText()

import 'dotenv/config';
import {Pipe} from 'langbase';

// 1. Initiate the Pipe.
const pipe = new Pipe({
	// Make sure you have a .env file with any pipe you wanna use.
	// As a demo we're using a pipe that has less wordy responses.
	apiKey: process.env.LANGBASE_PIPE_API_KEY!,
});

// 2. Generate a stream by asking a question
const stream = await pipe.streamText({
	messages: [{role: 'user', content: 'Who is an AI Engineer?'}],
});

// 3. Print the stream
for await (const chunk of stream) {
	// Streaming text part — a single word or several.
	const textPart = chunk.choices[0]?.delta?.content || '';

	// Demo: Print the stream — you can use it however.
	process.stdout.write(textPart);
}

Check out more examples in the docs →