AI Client Integration Guide: Multi-Platform Setup for NextChat, Open WebUI & ChatBox

Why Do You Need an AI Client?

Most people first experience AI chat through a web browser. But as you use it more, you will notice the limitations: poor chat history management, no way to compare multiple models side by side, and a basic interface with few features.

AI clients solve these problems. They act as a universal remote control that connects to different AI models (GPT-4o, Claude, DeepSeek, etc.), letting you switch freely within one app while managing conversation history and saving your favorite prompts.

This guide will walk you through setting up three popular AI clients from scratch, taking your AI experience to the next level.

Before You Start

You need two things to get started:

1. An AI platform account (such as Ciyuano) to obtain an API key
2. An AI client application (detailed below)

If you do not have a Ciyuano account yet, register on the official website. After registration, go to the "API Keys" page in the dashboard, click "Create New Key", and copy the generated key. Note: the key is only shown once, so save it securely.

3 Steps to Configure Your Client

Regardless of which client you choose, the configuration process is essentially the same — just three core steps:

Step 1: Get Your API Key

Log in to the Ciyuano dashboard and navigate to the "API Keys" management page. Click "Create New Key", give it a name (e.g., "NextChat dedicated"), and confirm. The system will generate a key string starting with sk-. Copy and save it immediately to a secure location.

Step 2: Fill in Connection Details

Open your chosen client app and find the "Settings" or "API Configuration" section. Enter the following information:

• API Endpoint: https://api.ciyuano.com/v1
• API Key: paste the key you copied earlier
• Model Name: choose based on your needs, such as gpt-4o, claude-4, or deepseek-chat

Step 3: Test and Start Using

After configuration, send a test message (e.g., "Hello, please introduce yourself"). If you receive a normal reply, the configuration is successful. You can now chat freely in the client.

Detailed Comparison of Three Clients

There are many AI clients available. Here are three of the most practical ones, covering mobile, desktop, and self-hosted scenarios.

NextChat — Best for Mobile

NextChat is one of the most popular AI chat clients, supporting both iOS and Android. Its advantages include:

• Multi-model switching: quickly switch between GPT-4o, Claude, DeepSeek, and more in one interface
• Conversation management: groups, search, and chat history export
• Prompt templates: built-in library of practical prompt templates for one-click use
• Plugin ecosystem: support for image generation, voice input, and other extensions

Setup: open NextChat, go to Settings, API Configuration, select "Custom API", enter the API endpoint and key, then save.

Open WebUI — Best for Self-Hosting

Open WebUI is an open-source AI web interface that can be deployed on your own server. Ideal for technical users who value data privacy:

• Docker one-click deployment: a single command starts the service
• RAG knowledge base: upload documents for AI to reference when answering questions
• Multi-user management: assign accounts to team members
• Web search plugin: AI can search the internet in real time

Setup: after deploying Open WebUI, go to Admin Settings, Connections, add the Ciyuano API endpoint and key, then save. All users can use this configuration.

ChatBox — Best for Desktop

ChatBox is a native desktop application supporting Windows, Mac, and Linux. Its key features:

• Offline capable: chat history saved locally, accessible without internet
• Local model support: connect to locally deployed models via Ollama
• Data export: export conversations as Markdown or JSON
• Clean interface: minimal design focused on conversation

Setup: download and install ChatBox, open Settings, select "OpenAI API" as the API type, enter the API endpoint and key, choose a default model, then save.

Common Issues and Solutions

Connection Failed?

• Check that the API endpoint is correct (no extra spaces or missing slashes)
• Confirm the API key is valid (you can regenerate one in the dashboard)
• Verify your network connection (some regions may require a proxy)

Empty Model List?

• Some clients require you to manually enter the model name instead of selecting from a list
• Confirm your account has access to the corresponding model

Slow Response?

• This is normal — different models have different response speeds
• GPT-4o and Claude typically respond within 2-5 seconds
• DeepSeek is faster, usually 1-3 seconds

Advanced Tips

After basic configuration, you can further enhance your experience:

• Create multiple API keys: use different keys for different clients for easier management and billing
• Set a default model: configure your most-used model as the default to save time
• Import prompt templates: find useful prompt templates online and import them into your client
• Enable plugins: if supported, turn on image generation, web search, and other extensions

Summary

AI clients are essential tools for improving your AI workflow. This guide covered three mainstream clients: NextChat for mobile users, Open WebUI for self-hosted scenarios, and ChatBox for desktop work.

Regardless of which you choose, the core setup is the same: get a key, fill in connection details, and test. Once configured, you will have an AI workstation that lets you switch models freely and manage conversations efficiently. Start today and leave the limitations of web-based AI behind.