Getting Started

Welcome to the CG SDK! This guide will help you get started with the CognipeerAI Gateway SDK.

What is CG SDK?

The CG SDK is the official TypeScript/JavaScript client library for CognipeerAI Gateway, a multi-tenant SaaS platform for AI and Agentic services. It provides:

• Chat Completions: OpenAI-compatible chat API with streaming
• Embeddings: Text vectorization for semantic search
• Vector Operations: Unified API for vector databases
• File Management: Upload and process documents
• Agent Tracing: Observability for agent executions

Prerequisites

• Node.js 18 or higher
• A CognipeerAI Gateway account and API key

Installation

Install the SDK using your preferred package manager:

npm

npm install @cognipeer/cgate-sdk

yarn

yarn add @cognipeer/cgate-sdk

pnpm

pnpm add @cognipeer/cgate-sdk

Get Your API Key

1. Sign up at CognipeerAI Gateway
2. Navigate to your dashboard
3. Go to Settings > API Tokens
4. Create a new API token
5. Copy the token (you'll need it in the next step)

Initialize the Client

Create a new client instance with your API key:

import { CGateClient } from '@cognipeer/cgate-sdk';

const client = new CGateClient({
  apiKey: 'your-api-key-here',
});

Environment Variables

For security, we recommend storing your API key in an environment variable:

const client = new CGateClient({
  apiKey: process.env.CGATE_API_KEY!,
});

Your First Request

Let's make a simple chat completion request:

const response = await client.chat.completions.create({
  model: 'gpt-4',
  messages: [
    {
      role: 'system',
      content: 'You are a helpful assistant.',
    },
    {
      role: 'user',
      content: 'What is the capital of France?',
    },
  ],
});

console.log(response.choices[0].message.content);
// Output: "The capital of France is Paris."

Streaming Example

Stream responses for real-time output:

const stream = await client.chat.completions.create({
  model: 'gpt-4',
  messages: [
    {
      role: 'user',
      content: 'Tell me a short story about a robot.',
    },
  ],
  stream: true,
});

for await (const chunk of stream) {
  const content = chunk.choices[0]?.delta?.content;
  if (content) {
    process.stdout.write(content);
  }
}

Configuration Options

The client supports several configuration options:

const client = new CGateClient({
  apiKey: 'your-api-key',
  baseURL: 'https://api.cognipeer.com', // Optional: custom API URL
  timeout: 60000, // Optional: request timeout (ms)
  maxRetries: 3, // Optional: max retry attempts
  fetch: customFetch, // Optional: custom fetch implementation
});

Next Steps

Now that you've set up the SDK, explore these topics:

• Authentication - Learn about API keys and security
• Configuration - Advanced client configuration
• Chat API - Detailed chat completions documentation
• Examples - Real-world usage examples

TypeScript Support

The SDK is written in TypeScript and includes full type definitions:

import type {
  ChatCompletionRequest,
  ChatCompletionResponse,
  EmbeddingRequest,
} from '@cognipeer/cgate-sdk';

// Full intellisense and type checking
const request: ChatCompletionRequest = {
  model: 'gpt-4',
  messages: [
    { role: 'user', content: 'Hello!' }
  ],
};

Getting Help

• 📖 Full Documentation
• 🐛 Report Issues
• 📧 Email Support