Aisool SDK

The official JavaScript / TypeScript SDK for interacting with Aisool AI – Sool Max v2.

Aisool SDK provides a simple and secure way to communicate with the Aisool AI API using the OpenAI-compatible Chat Completions and Image Generations format.

✨ Features

• 🚀 Simple and lightweight
• 🔐 Secure Bearer token authentication
• 💬 Chat-based AI interaction (Sool Max v2)
• 🎨 AI Image Generation (Powered by FreeGen)
• 📦 Full TypeScript support
• ⚡ Works in Node.js and modern frameworks (Next.js, Vite, etc.)
• 🔄 OpenAI-compatible API format

📦 Installation

Install via your favorite package manager:

npm install aisool-sdk
# or
pnpm add aisool-sdk
# or
yarn add aisool-sdk

🚀 Quick Start

Chat Completion

import { AisoolClient } from 'aisool-sdk';

const client = new AisoolClient('YOUR_API_KEY');

async function askAI() {
  try {
    const response = await client.chat('Hello! How are you?');
    console.log('AI Response:', response.content);
  } catch (error: any) {
    console.error('Error:', error.message);
  }
}

askAI();

Image Generation

async function generateArt() {
  try {
    const response = await client.generateImage('A futuristic neon city at night');
    console.log('Image URL (Base64):', response.data[0].url);
  } catch (error: any) {
    console.error('Error:', error.message);
  }
}

generateArt();

🔑 Authentication

You must provide your Aisool API Key when creating the client.

const client = new AisoolClient('YOUR_API_KEY');

⚠️ Never expose your API key in frontend public code. Use environment variables: const client = new AisoolClient(process.env.AISOOL_API_KEY!);

🎨 Image Generation Method

client.generateImage(prompt: string)

Generates a high-quality image based on the provided text prompt. Note that this method uses Selenium-based automation on the backend, so it may take up to 30–60 seconds.

Response Format:

interface ImageResponse {
  created: number;
  data: Array<{
    url: string; // Base64 encoded image string or URL
  }>;
}

💬 Chat Method

client.chat(message: string)

Sends a message to Sool Max v2 (sool-max model) and returns a structured AI response.

Response Format:

interface ChatResponse {
  role: 'assistant' | 'user';
  content: string;
  finish_reason: string | null;
  usage?: {
    prompt_tokens: number;
    completion_tokens: number;
    total_tokens: number;
  };
}

🛠 Error Handling

Always wrap requests in try/catch. If you call a chat completion without the correct internal model mapping, the SDK will catch the error returned by the server.

try {
  const res = await client.chat('Hello AI');
} catch (error: any) {
  // Catch authentication errors, model mismatch (400), or timeouts (504)
  console.error(error.message);
}

🌍 Environment Support

• Node.js 18+
• Next.js / Nuxt.js
• Vite / React / Vue
• TypeScript projects
• Any environment supporting fetch

📄 License

MIT License © 2026 AiSool

🔗 Official Links

• Website: https://aisool.com
• API Dashboard: https://aisool.com/api

Built with ❤️ for developers.