Q-Ollama

A Node.js package for interacting with Ollama and Baichuan AI models with flexible API and CLI support.

Features

• 🤖 Support for Ollama and Baichuan AI models
• 🔄 Dynamic provider switching
• 💬 Interactive chat and single message support
• 🔧 Full TypeScript support
• 🚀 Command-line tool support
• 🐛 Debug mode with detailed logging
• 📚 Comprehensive test cases

Installation

npm install q-ollama

Quick Start

1. Using Ollama

Make sure Ollama service is running (default: http://localhost:11434)

const { QOllama, ProviderType } = require('q-ollama');

// Create instance
const qollama = new QOllama({
  provider: ProviderType.OLLAMA,
  ollamaBaseUrl: 'http://localhost:11434',
  defaultModel: 'qwen3:8b',
  debug: true
});

// Quick chat
async function chat() {
  const response = await qollama.quickChat('Hello, please introduce yourself');
  console.log('AI Response:', response.content);
}

chat();

2. Using Baichuan Model

Set environment variable BAICHUAN_API_KEY or provide API key directly:

const { QOllama, ProviderType } = require('q-ollama');

const qollama = new QOllama({
  provider: ProviderType.BAICHUAN,
  baichuanApiKey: 'your-api-key-here', // or use environment variable
  defaultModel: 'Baichuan2-Turbo'
});

async function chat() {
  const response = await qollama.quickChat('Hello');
  console.log('Baichuan Response:', response.content);
}

chat();

3. Dynamic Provider Switching

// Start with Ollama
const qollama = new QOllama({
  provider: ProviderType.OLLAMA,
  defaultModel: 'qwen3:8b'
});

console.log('Current provider:', qollama.getCurrentProvider()); // ollama

// Switch to Baichuan
qollama.switchProvider({
  provider: ProviderType.BAICHUAN,
  baichuanApiKey: process.env.BAICHUAN_API_KEY
});

console.log('Switched provider:', qollama.getCurrentProvider()); // baichuan

API Reference

QOllama Class

Constructor

new QOllama(config: QOllamaConfig)

Configuration options:

interface QOllamaConfig {
  provider: ProviderType;          // Model provider
  ollamaBaseUrl?: string;          // Ollama service URL
  baichuanApiKey?: string;         // Baichuan API key
  defaultModel?: string;           // Default model
  debug?: boolean;                 // Debug mode
}

Methods

• chat(messages: ChatMessage[], options?: ChatOptions): Promise<ChatResponse> - Send chat messages
• quickChat(prompt: string, options?: ChatOptions): Promise<ChatResponse> - Quick single message
• switchProvider(newConfig: QOllamaConfig): void - Switch model provider
• getCurrentProvider(): string - Get current provider
• supportsStreaming(): boolean - Check if streaming is supported
• listModels(): Promise<string[]> - List available models
• setDebug(debug: boolean): void - Set debug mode

Helper Functions

const { createQOllama, createOllamaProvider, createBaichuanProvider } = require('q-ollama');

// Quick instance creation
const qollama1 = createQOllama(config);
const qollama2 = createOllamaProvider('http://localhost:11434', true);
const qollama3 = createBaichuanProvider('your-api-key', true);

Command Line Tool

After installation, use the q-ollama command:

Interactive Chat

# Using Ollama
q-ollama chat --provider ollama --model qwen3:8b

# Using Baichuan
q-ollama chat --provider baichuan --model Baichuan2-Turbo --key YOUR_API_KEY

Single Message

q-ollama message "Hello world" --provider ollama --model qwen3:8b

List Available Models

q-ollama list-models --provider ollama

Full Command Help

q-ollama --help

Debug Mode

Enable debug mode to see detailed request and response information:

const qollama = new QOllama({
  provider: ProviderType.OLLAMA,
  debug: true  // Enable debug
});

// Or enable at runtime
qollama.setDebug(true);

Debug output includes:

• Method call parameters
• API request details
• Response data
• Error information

Environment Variables

• BAICHUAN_API_KEY - Baichuan model API key

Development

Build Project

npm run build

Run Tests

npm test

Development Mode

npm run dev

Examples

Check the examples/ directory for complete examples:

node examples/basic-usage.js

License

MIT

Contributing

Issues and Pull Requests are welcome!

Support

If you encounter issues:

1. Check debug mode output
2. Ensure related services are running properly
3. Check test cases for correct usage