@jutech-devs/universal-ai

Universal AI SDK

A unified SDK for interacting with ALL major AI providers with React hooks support.

Supported Providers & Models

🤖 OpenAI

• Latest: gpt-4o, gpt-4o-mini, o1-preview, o1-mini
• GPT-4: gpt-4-turbo, gpt-4, gpt-4-32k
• GPT-3.5: gpt-3.5-turbo, gpt-3.5-turbo-16k
• Legacy: text-davinci-003, code-davinci-002

🧠 Anthropic Claude

• Claude 3.5: claude-3-5-sonnet-20241022, claude-3-5-haiku-20241022
• Claude 3: claude-3-opus-20240229, claude-3-sonnet-20240229, claude-3-haiku-20240307
• Claude 2: claude-2.1, claude-2.0, claude-instant-1.2

🔍 Google Gemini

• Gemini 2.0: gemini-2.0-flash-exp
• Gemini 1.5: gemini-1.5-pro-002, gemini-1.5-pro, gemini-1.5-flash-002, gemini-1.5-flash, gemini-1.5-flash-8b
• Legacy: gemini-pro, gemini-pro-vision, gemini-ultra

🚀 Groq (Ultra-Fast Inference)

• Llama 3.3: llama-3.3-70b-versatile
• Llama 3.2: llama-3.2-90b-text-preview, llama-3.2-11b-text-preview
• Llama 3.1: llama-3.1-70b-versatile, llama-3.1-8b-instant
• Mixtral: mixtral-8x7b-32768
• Gemma: gemma2-9b-it, gemma-7b-it

🎯 Cohere

• Command R: command-r-plus, command-r
• Command: command, command-nightly, command-light

🌟 Mistral AI

• Large: mistral-large-latest, mistral-large-2407
• Medium/Small: mistral-medium-latest, mistral-small-latest
• Open Source: open-mistral-7b, open-mixtral-8x7b, open-mixtral-8x22b
• Code: codestral-latest, codestral-2405

🤖 xAI (Grok)

• Grok: grok-beta, grok-vision-beta

🧮 DeepSeek

• Models: deepseek-chat, deepseek-coder, deepseek-reasoner

API Reference

UniversalAI Class

const ai = new UniversalAI(config);

// Chat completion
const response = await ai.chat(messages);

// Streaming
for await (const chunk of ai.stream(messages)) {
  console.log(chunk.content);
}

// Get available models
const models = ai.getModels();

// Update configuration
ai.updateConfig({ temperature: 0.8 });

React Hooks

useAI

const {
  messages,      // Array of chat messages
  isLoading,     // Loading state
  error,         // Error message
  response,      // Last AI response
  sendMessage,   // Function to send a message
  clearMessages, // Function to clear chat history
  updateConfig,  // Function to update AI config
  models,        // Available models for current provider
} = useAI(config);

useAIStream

const {
  messages,         // Array of chat messages
  streamingContent, // Current streaming content
  isLoading,        // Loading state
  error,            // Error message
  sendMessage,      // Function to send a message
  clearMessages,    // Function to clear chat history
  models,           // Available models
} = useAIStream(config);

Configuration Options

interface UniversalAIConfig {
  provider: 'openai' | 'anthropic' | 'google';
  apiKey: string;
  model: string;
  temperature?: number;    // 0-1, default: 0.7
  maxTokens?: number;      // Max response tokens
  topP?: number;           // 0-1, nucleus sampling
  baseURL?: string;        // Custom API endpoint
}

Examples

Multi-Provider Chat

import { UniversalAI } from '@jutech-devs/universal-ai';

// OpenAI
const openai = new UniversalAI({
  provider: 'openai',
  apiKey: process.env.OPENAI_API_KEY!,
  model: 'gpt-4',
});

// Anthropic
const claude = new UniversalAI({
  provider: 'anthropic',
  apiKey: process.env.ANTHROPIC_API_KEY!,
  model: 'claude-3-5-sonnet-20241022',
});

// Google
const gemini = new UniversalAI({
  provider: 'google',
  apiKey: process.env.GOOGLE_API_KEY!,
  model: 'gemini-1.5-pro',
});

const messages = [
  { role: 'user', content: 'Explain quantum computing' }
];

const [openaiResponse, claudeResponse, geminiResponse] = await Promise.all([
  openai.chat(messages),
  claude.chat(messages),
  gemini.chat(messages),
]);

Custom System Prompts

import { createSystemMessage, createUserMessage } from '@jutech-devs/universal-ai';

const messages = [
  createSystemMessage('You are a helpful coding assistant.'),
  createUserMessage('How do I create a React component?'),
];

const response = await ai.chat(messages);

Token Usage Tracking

const response = await ai.chat(messages);

if (response.usage) {
  console.log(`Tokens used: ${response.usage.totalTokens}`);
  console.log(`Prompt tokens: ${response.usage.promptTokens}`);
  console.log(`Completion tokens: ${response.usage.completionTokens}`);
}

License

MIT © JuTech Devs