LangChain

Use LangChain with Codzen to access multiple AI models through a unified interface with provider failover, budget controls, and usage visibility.

Why Use Codzen with LangChain?

LangChain provides a standard interface for working with chat models. By connecting LangChain to Codzen, you get:

• Model Flexibility: Access to multiple AI providers and models through a single API
• Provider Failover: Automatic fallback to alternative providers when one is unavailable
• Budget Controls: Set spending limits and monitor usage across your organization
• Usage Visibility: Track all API calls and costs in your Codzen dashboard

Quick Start

This guide will get you running LangChain with Codzen in just a few minutes.

Prerequisites

• A Codzen API key
• LangChain installed in your project

TypeScript

Install the required packages:

npm install @langchain/openai @langchain/core

Configure LangChain to use Codzen:

import { HumanMessage } from '@langchain/core/messages';
import { ChatOpenAI } from '@langchain/openai';

const chat = new ChatOpenAI(
  configuration: {
    baseURL: 'https://codzen.ai/v1',
    apiKey: '<CODZEN_API_KEY>',
  },
  model: 'gpt-5-mini',
);

// Example usage
const response = await model.invoke([
  new HumanMessage('What is the meaning of life?'),
]);

console.log('Response:', response.content);

Python

Install the required packages:

pip install pydantic langchain-openai

Configure LangChain to use Codzen:

from langchain_openai import ChatOpenAI
from pydantic import SecretStr

model = ChatOpenAI(
    base_url="https://codzen.ai/v1",
    api_key=SecretStr("<CODZEN_API_KEY>"),
    model="gpt-5-mini",
)

# Example usage
response = model.invoke("What is the meaning of life?")
print(response.content)

Configuration Options

Available Models

Codzen provides access to models from multiple providers. Visit the Models page to see all available models and their capabilities.

Resources

• LangChain Python Documentation
• LangChain JavaScript Documentation
• Codzen API Reference

Troubleshooting

• Model Not Found: Check that the model identifier is correct.
• Rate Limits: If you encounter rate limits, consider implementing retry logic or contact support to increase your limits.