AIOS Platform Documentation

Welcome to the AIOS Platform documentation. This guide will help you understand how to build, deploy, and manage AI agents using our platform.

AIOS Platform provides a comprehensive suite of tools and services for AI agent development, including multiple LLM support, smart LLM routing, agent memory service, and RAG (Retrieval-Augmented Generation) capabilities.

What is AIOS Platform?

AIOS (AI Agent Operating System) Platform is a cloud service that provides developers and enterprises with cloud-based agent hosting services. Developers can build agents based on AIOS APIs and deploy them to reach millions of users.

Key features include:

  • Multiple LLM support (GPT-4, Claude, Llama, etc.)
  • Smart LLM routing for optimal performance and cost efficiency
  • Agent memory service for persistent context
  • RAG service for enhanced knowledge retrieval
  • One-click deployment and scaling

Quick Start

Get started with AIOS Platform in minutes. Follow these steps to create and deploy your first AI agent.

1. Install the AIOS SDK

# Using npm
npm install @aios/sdk

# Using yarn
yarn add @aios/sdk

# Using pip
pip install aios-sdk

2. Initialize Your Agent

// JavaScript/TypeScript
import { Agent } from '@aios/sdk';

const agent = new Agent({
  name: 'My First Agent',
  description: 'A simple AI assistant',
  apiKey: 'your-api-key',
  defaultModel: 'gpt-4'
});

3. Define Agent Behavior

// Define how your agent responds to user input
agent.onMessage(async (message, context) => {
  const response = await agent.llm.generate({
    prompt: message.content,
    systemPrompt: 'You are a helpful assistant.',
    maxTokens: 500
  });
  
  return {
    content: response.text
  };
});

4. Deploy Your Agent

// Deploy to AIOS Platform
const deployment = await agent.deploy();

console.log(`Agent deployed: ${deployment.url}`);

Authentication

To use the AIOS Platform API, you need to authenticate your requests using an API key. You can generate an API key from your AIOS Platform dashboard.

API Key Authentication

Include your API key in the request headers:

// JavaScript example
const response = await fetch('https://api.aiosplatform.com/v1/agents', {
  headers: {
    'Authorization': 'Bearer YOUR_API_KEY',
    'Content-Type': 'application/json'
  }
});
# cURL example
curl -X GET \
  'https://api.aiosplatform.com/v1/agents' \
  -H 'Authorization: Bearer YOUR_API_KEY' \
  -H 'Content-Type: application/json'

Agent API

The Agent API allows you to create, manage, and deploy AI agents on the AIOS Platform.

Create an Agent

POST
/v1/agents

Create a new AI agent with the specified configuration.

Request Parameters

Parameter Type Description
name Required string The name of the agent
description Optional string A description of the agent's purpose and capabilities
defaultModel Required string The default LLM model to use (e.g., "gpt-4", "claude-2")
systemPrompt Optional string The system prompt that defines the agent's behavior

Example Request

POST /v1/agents
Content-Type: application/json
Authorization: Bearer YOUR_API_KEY

{
  "name": "Customer Support Agent",
  "description": "An agent that handles customer support inquiries",
  "defaultModel": "gpt-4",
  "systemPrompt": "You are a helpful customer support agent for ACME Inc."
}

Example Response

{
  "id": "agt_123456789",
  "name": "Customer Support Agent",
  "description": "An agent that handles customer support inquiries",
  "defaultModel": "gpt-4",
  "systemPrompt": "You are a helpful customer support agent for ACME Inc.",
  "createdAt": "2025-05-04T12:34:56Z",
  "status": "created"
}

Model API

The Model API allows you to interact with various LLM models supported by AIOS Platform.

Generate Text

POST
/v1/models/{model}/generate

Generate text using the specified LLM model.

Request Parameters

Parameter Type Description
prompt Required string The prompt to generate text from
systemPrompt Optional string System instructions for the model
maxTokens Optional integer Maximum number of tokens to generate (default: 1000)
temperature Optional float Controls randomness (0-1, default: 0.7)