qwen-api

Unofficial Qwen Chat API client for Node.js and TypeScript.

Install

To use this package, you need to install the dependencies by running the following command:

npm install qwen-api

Setup

You need to set up the API token and other configurations by storing them in a .env file in the root of your project directory. Example of a .env file:

QWEN_AUTH_TOKEN=your_auth_token_here
QWEN_COOKIE=your_cookie_here

Make sure to replace your_auth_token_here and your_cookie_here with your valid authentication token and cookie.

Usage Examples

1. Using the Streaming API

In streaming mode, you can receive data incrementally as the request is being processed.

import { QwenClient } from 'qwen-api';
import 'dotenv/config';

async function main() {
    const client = new QwenClient();
    const stream = await client.completion.create({
        messages: [{
            role: "user",
            content: "What is the capital of Indonesia?",
            web_search: false,
            thinking: false
        }],
        model: "qwen-max-latest",
        stream: true,
    });

    console.log(`User: What is the capital of Indonesia?`);
    console.log(`Assistant:`);
    if (stream instanceof ReadableStream) {
        for await (const chunk of stream) {
            process.stdout.write(chunk.delta?.content || '');
        }
    } else {
        console.error('Stream is not a ReadableStream');
    }
}

main().catch(console.error);

2. Using the Non-Streaming API

If you prefer to get the response directly without streaming, you can use the non-streaming mode as shown below:

import { QwenClient } from 'qwen-api';
import 'dotenv/config';

async function main() {
    const client = new QwenClient();
    const response = await client.completion.create({
        messages: [{
            role: "user",
            content: "What is the capital of Indonesia?",
            web_search: false,
            thinking: false
        }],
        model: "qwen-max-latest",
        stream: false,
    });

    console.log(response);
}

main().catch(console.error);

API Reference

QwenClient

The main class to interact with the Qwen API.

QwenClient.completion.create(options: CompletionOptions)

This method sends a request to the Qwen API.

• options: The request configuration object.

  • messages: An array of messages to send.

    • role: The role of the message sender, either "user" or "assistant".
    • content: The content of the message.
    • web_search: Whether to perform a web search (default: false).
    • thinking: Whether the model is processing its thoughts (default: false).

  • model: The name of the model being used, e.g., "qwen-max-latest".

  • stream: true for streaming, false for immediate response.

Example Response

For streaming mode, data is received incrementally through a ReadableStream. If not using streaming, the response will be in the form of a JSON object containing the result of the request.

Testing

You can run the tests provided to verify everything is working as expected:

npm run test:client
npm run test:stream

Make sure your .env file is configured correctly with a valid API key.