Hexa Generative AI - Node.js SDK

Official Node.js/TypeScript client for Hexa AI API.

Installation

npm install hexgenerative-ai
# or
yarn add hexgenerative-ai
# or
pnpm add hexgenerative-ai

Quick Start

import HexaAI from "hexgenerative-ai";

const client = new HexaAI({
  apiKey: "hgx-your-api-key",
});

async function main() {
  const response = await client.chat.completions.create({
    model: "hexa-pro",
    messages: [
      { role: "user", content: "Explain quantum computing in simple terms" },
    ],
  });

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

main();

Available Models

| Model | Description | Best For | |-------|-------------|----------| | hexa-instant | Fastest model | Quick responses | | hexa-balanced | General purpose | Most use cases | | hexa-reasoning | Deep analysis | Complex reasoning | | hexa-advanced | Coding expert | Programming | | hexa-pro | Premium quality | Best results | | hexa-vision-scout | Vision model | Images |

Smart Routing

// By task type
const response = await client.chat.completions.create({
  task: "coding",
  messages: [{ role: "user", content: "Write a function" }],
});

// By optimization
const response = await client.chat.completions.create({
  optimize_for: "speed",
  messages: [{ role: "user", content: "Quick answer" }],
});

// Auto-select
const response = await client.chat.completions.create({
  auto_select: true,
  messages: [{ role: "user", content: "Your message" }],
});

Error Handling

import HexaAI, { HexaAIError } from "hexgenerative-ai";

const client = new HexaAI({ apiKey: "hgx-your-key" });

try {
  const response = await client.chat.completions.create({
    model: "hexa-pro",
    messages: [{ role: "user", content: "Hello" }],
  });
} catch (error) {
  if (error instanceof HexaAIError) {
    console.log("Error:", error.message);
    console.log("Status:", error.statusCode);
  }
}

TypeScript Support

Full TypeScript support with type definitions included.

import HexaAI, { ChatMessage, ChatCompletion } from "hexgenerative-ai";

const messages: ChatMessage[] = [
  { role: "system", content: "You are a helpful assistant" },
  { role: "user", content: "Hello!" },
];

const response: ChatCompletion = await client.chat.completions.create({
  model: "hexa-pro",
  messages,
});

Agentic Features

Agent Tasks

const result = await client.agent.run({
  task: "Research AI trends and summarize",
  model: "hexa-ultra",
});
console.log(result);

RAG (Knowledge Base)

// Upload document
await client.rag.upload({
  title: "Company Policy",
  content: "Employees get 30 days leave...",
});

// Search
const results = await client.rag.search({ query: "leave policy" });

Context Management (300K Tokens)

const session = await client.context.create({
  system_prompt: "You are a helpful assistant",
});

await client.context.add(session.data.session_id, {
  role: "user",
  content: "Hello!",
});

Code Execution

const result = await client.code.execute({
  code: "print(sum(range(100)))",
});
console.log(result.data.output);

Tools

const tools = await client.tools.list();
console.log(`Found ${tools.data.count} tools`);

License

MIT