Quick Start

This guide will help you make your first API call with the Exnest AI SDK.

Prerequisites

Before you begin, make sure you have:

• An Exnest AI API Key (Get one from the Exnest Dashboard)
• Node.js 18+ or Python 3.8+ installed

Making Your First Request

Node.js

Using the Advanced Client (Recommended)

The ExnestAI client provides full control over configuration, retries, and access to all features including EBC.

import { ExnestAI } from '@exnest-dev/ai';

const exnest = new ExnestAI({
  apiKey: 'your-api-key',
  baseUrl: 'https://api.exnest.app/v1', // Optional
  timeout: 30000, // Optional
  retries: 3, // Optional
  retryDelay: 1000, // Optional
  debug: true // Optional
});

// Simple chat completion
const response = await exnest.chat('gpt-4.1-mini', [
  { role: 'user', content: 'Hello, how are you?' }
]);

console.log(response.data.choices[0].message.content);

Using the Simple Wrapper

For quick and simple use cases.

import { ExnestWrapper } from '@exnest-dev/ai';

const exnest = new ExnestWrapper('your-api-key');

// Simple response
const response = await exnest.response('claude-3-haiku', 'What is TypeScript?');
console.log(response.data.choices[0].message.content);

Python

Using the Advanced Client (Recommended)

For full control over configuration, retries, timeouts, and access to EBC (Exnest Brain Core) features.

import asyncio
import os
from exnestai import ExnestAI, ExnestMessage

async def main():
    api_key = os.getenv("EXNEST_API_KEY")

    # Initialize the advanced client
    client = ExnestAI(
        api_key=api_key,
        timeout=60000,  # 60 seconds
        retries=2,
        debug=True
    )

    # --- Standard Chat ---
    chat_response = await client.chat(
        model="openai:gpt-4o-mini",
        messages=[ExnestMessage(role="user", content="Hello! What can you tell me about ExnestAI?")],
        exnest_metadata=True
    )
    if not chat_response.error:
        print(f"Chat Response: {chat_response.choices[0].message.content}")

if __name__ == "__main__":
    asyncio.run(main())

Using the Simple Wrapper

For straightforward use cases.

import asyncio
import os
from exnestai import ExnestWrapper, ExnestMessage

async def main():
    # Initialize the wrapper
    api_key = os.getenv("EXNEST_API_KEY")
    exnest = ExnestWrapper(api_key=api_key)

    # Perform a chat completion
    chat_response = await exnest.chat(
        model="openai:gpt-4o-mini",
        messages=[ExnestMessage(role="user", content="Hello, how are you?")]
    )
    if not chat_response.error:
        print(f"Wrapper Chat Response: {chat_response.choices[0].message.content}")

if __name__ == "__main__":
    asyncio.run(main())

Compatibility Layer

Using LangChain (or other OpenAI-compatible clients)

You can use standard OpenAI-compatible clients by pointing the baseURL to Exnest’s API.

import { ChatOpenAI } from "@langchain/openai";

export function createLLM(options: AIServiceOptions = {}) {
    const apiKey = config.exnestApiKey;

    if (!apiKey) {
        throw new Error("EXNEST_API_KEY is not configured");
    }

    return new ChatOpenAI({
        apiKey,
        temperature: options.temperature ?? 0.7,
        streaming: options.streaming ?? false,
        configuration: {
            baseURL: "https://api.exnest.app/v1",
        },
        modelName: options.modelName ?? "gpt-4.1-mini",
    });
}