Skip to content

XPULINK AI Model API Quick Start Guide

This guide will help you quickly get started with the XPULINK AI platform's cloud-based model API. We'll use Python and the requests library to call the qwen3-32b model.

Prerequisites

  • Python 3.7+
  • requests library (pip install requests)
  • XPULINK AI platform account and API Key

Configuration Steps

1. Obtain API Key

  1. Log in to XPULINK AI platform (https://www.xpulink.ai)
  2. Get your API Key from the user center
  3. Set the API Key as an environment variable

2. Set Environment Variable

# Linux/Mac
export XPULINK_API_KEY="your_api_key_here"

# Windows
set XPULINK_API_KEY=your_api_key_here

3. Model Information

Complete Code Example

The following code demonstrates how to call XPULINK AI's cloud-based model for conversation:

import os
import requests

# Read API Key from environment variable
API_KEY = os.getenv("XPULINK_API_KEY")
if not API_KEY:
    raise ValueError("Please set XPULINK_API_KEY in environment variables")

# Cloud model API information
MODEL_NAME = "qwen3-32b"
BASE_URL = "https://www.xpulink.ai/v1/chat/completions"

# Construct request headers
headers = {
    "Authorization": f"Bearer {API_KEY}",
    "Content-Type": "application/json"
}

# Construct request body
payload = {
    "model": MODEL_NAME,
    "messages": [
        {"role": "user", "content": "Hello, please briefly introduce yourself."}
    ],
    "max_tokens": 50,
    "temperature": 0.7
}

# Send request and print result
try:
    response = requests.post(BASE_URL, headers=headers, json=payload, timeout=30)
    response.raise_for_status()
    result = response.json()
    print("Model response:", result["choices"][0]["message"]["content"])
    print("Test passed! Cloud model is working correctly.")
except Exception as e:
    print("Test failed:", e)

Parameter Description

Request Parameters

  • model: Model name, using "qwen3-32b" here
  • messages: Array of conversation messages, containing role and content
  • max_tokens: Maximum number of tokens to generate, controls response length
  • temperature: Temperature parameter, controls randomness (0-1), higher values mean more random

Response Format

A successful response returns JSON formatted data with main fields including: - choices[0].message.content: Model-generated response content - usage: Token usage statistics - model: Name of the model used

Running the Test

Save the code as test_xpulink.py and run:

python test_xpulink.py

If configured correctly, you will see the model's self-introduction response and a success message.

Troubleshooting

Common Errors

  1. API Key Error: Ensure environment variable is set correctly
  2. Network Timeout: Check network connection, or increase timeout value
  3. Permission Error: Confirm API Key is valid and has sufficient quota

Debugging Tips

  • Print complete error messages for more details
  • Use curl command to test API connectivity
  • Check XPULINK platform status page

Extended Usage

Based on this basic example, you can: - Modify messages content for multi-turn conversations - Adjust temperature parameter to control response style - Add system role messages to set system prompts - Include additional request parameters like top_p, frequency_penalty, etc.