Prerequisites
Before you begin, make sure:
ChatBox is installed
Download and install the version for your OS from ChatBox GitHub, or visit the ChatBox website.
You have a MaiToken API key
Sign in to the MaiToken Console to get your API key (it starts with
sk-).
Tip: If you don't have a MaiToken account yet, sign up at MaiToken first and get your API key.
Step 1: Launch ChatBox and Start Configuring
When launching ChatBox for the first time or adding a new AI provider:
- Launch the ChatBox app
- On first use, the configuration wizard opens automatically
- Click the "Use your own API Key or local model" button
- If you've configured it before, click the ⚙️ Settings icon at the bottom left, or use the shortcut
Ctrl+,(Windows/Linux) /Cmd+,(macOS)

Step 2: Configure the MaiToken API
2.1 Choose an AI Model Provider
On the settings page:
- Find the AI Model Provider or AI Provider setting
- From the dropdown, choose OpenAI API (for GPT-series models), Claude API (for Claude-series models), or Gemini API (for Gemini-series models)

2.2 Configure the API Details
After selecting OpenAI API, fill in the following:
| Setting | Value |
|---|---|
| API Key | Your MaiToken API key (sk-xxxxxxxxxxxx) |
| API Host or API Domain | https://api.maitoken.com |

After selecting Claude API, fill in the following:
| Setting | Value |
|---|---|
| API Key | Your MaiToken API key (sk-xxxxxxxxxxxx) |
| API Host or API Domain | https://api.maitoken.com/v1 |

After selecting Gemini API, fill in the following:
| Setting | Value |
|---|---|
| API Key | Your MaiToken API key (sk-xxxxxxxxxxxx) |
| API Host or API Domain | https://api.maitoken.com/ |

Important:
- For the OpenAI API, set the API Host to
https://api.maitoken.com(without the/v1suffix) - For the Claude API, set the API Host to
https://api.maitoken.com/v1(with the/v1suffix) - For the Gemini API, set the API Host to
https://api.maitoken.com/(without the/v1suffix) - The API Key must be a
sk--prefixed key obtained from the MaiToken Console - Make sure your API key has sufficient balance
2.3 Choose a Model
Once configured, select the model you want in the Model dropdown:
Recommended models:
| Model Name | Model ID | Highlights |
|---|---|---|
| GPT-5 | gpt-5 |
The latest, most powerful model |
| GPT-4o | gpt-4o or chatgpt-4o-latest |
High-quality conversation |
| GPT-4o Mini | gpt-4o-mini |
Fast and economical |
| Claude Sonnet 4.5 | claude-sonnet-4-5-20250929 |
Strong at code and reasoning |
| Claude Haiku 4.5 | claude-haiku-4-5-20251001 |
Fast responses |
| Gemini 2.0 Flash | gemini-2.0-flash-exp |
Multimodal support |
Special note for using Claude models: If you choose a Claude model (such as claude-sonnet-4-5-20250929 or claude-haiku-4-5-20251001), you need the following additional configuration:
- Switch to the Claude API provider:
- In settings, change the AI Model Provider to Claude API
- Keep the API Host as
https://api.maitoken.com/v1 - Use your MaiToken API key (
sk-xxxxxxxxxxxx) as the API Key
- Or add a custom request header (if using the OpenAI API provider):
- Some versions of ChatBox support adding custom headers
- In advanced settings, add the header:
anthropic-version: 2023-06-01
Recommended approach: Use OpenAI API for GPT-series models, Claude API for Claude-series models, and Gemini API for Gemini-series models.
Performance suggestions:
- 💰 Best value:
gpt-4o-mini,claude-haiku-4-5-20251001 - 🚀 High performance:
gpt-5,gpt-4o,claude-sonnet-4-5-20250929 - ⚡ Fast responses:
gemini-2.0-flash-exp,gpt-4o-mini
Step 3: Start Using It
Once configured, you can start chatting with the AI in ChatBox:
- Return to the main screen
- Type your question or request in the input box
- Press
Enteror click the send button - The AI generates a reply using MaiToken's models

Adjust Model Parameters (optional)
You can adjust the following parameters as needed:
| Parameter | Description | Recommended value |
|---|---|---|
| Temperature | Controls output randomness | 0.7 (creative) / 0.3 (precise) |
| Max Tokens | Maximum output length | 2000-4000 |
| Top P | Nucleus sampling parameter | 0.9 |
Advanced Features
Multi-Session Management
ChatBox supports creating multiple conversation sessions:
- Click the New conversation button
- Create separate sessions for different tasks
- Switch between conversations in the session list on the left
Save and Export Conversations
- Right-click a conversation session
- Choose the Export option
- Export as Markdown, JSON, and other formats
Use Prompt Templates
- Find the Prompts setting in the settings
- Save frequently used prompt templates
- Quickly invoke preset prompts in a conversation
FAQ
Q1: Can't connect to the MaiToken service?
Solution:
- Check the API Host:
- Make sure it is
https://api.maitoken.com - Don't add
/v1or any other path
- Make sure it is
- Verify the API Key:
- Confirm the API Key is correct and starts with
sk- - Check whether the key is valid in the MaiToken Console
- Confirm the API Key is correct and starts with
- Check your network connection:
- Make sure you can reach
https://api.maitoken.com - If you're in mainland China, you may need to configure a proxy
- Make sure you can reach
Q2: MaiToken's models don't show up in the model list?
Solution:
- Enter the model name manually:
- Type the model ID directly in the model input box
- For example:
gpt-4o,gpt-4o-mini,claude-sonnet-4-5-20250929
- Refresh the model list:
- Restart the ChatBox app
- Reconfigure the API details
Q3: Getting an error during a conversation?
Common errors and solutions:
| Error Message | Cause | Solution |
|---|---|---|
401 Unauthorized |
API Key invalid or expired | Get a new API Key and update the config |
429 Too Many Requests |
Request rate exceeded | Wait a moment and retry |
500 Internal Server Error |
Temporary server failure | Wait a few minutes and retry |
insufficient_quota |
Insufficient account balance | Top up in the Console |
Q4: How do I view API usage and cost?
Sign in to the MaiToken Console to view:
- 📊 API call statistics
- 💰 Cost breakdown
- 📈 Usage trend charts
- 🔍 Detailed request logs
Q5: Which platforms does ChatBox support?
ChatBox supports several platforms:
- 🪟 Windows - Windows 10/11
- 🍎 macOS - macOS 10.15+
- 🐧 Linux - major distributions
- 🌐 Web - browser version
Tips
1. Keyboard Shortcuts
Make full use of ChatBox's shortcuts to boost your efficiency:
| Shortcut | Function |
|---|---|
Ctrl/Cmd + Enter |
Send message |
Ctrl/Cmd + N |
New conversation |
Ctrl/Cmd + K |
Search conversations |
Ctrl/Cmd + , |
Open settings |
Ctrl/Cmd + / |
Show shortcut help |
2. Optimize Your Prompts
Write better prompts to get better replies:
❌ A poor prompt:
Help me write some code✅ A good prompt:
Please help me write a Python function that:1. Takes a list of strings as input2. Filters out strings shorter than 3 characters3. Returns the result sorted alphabeticallyPlease include detailed comments and example usage3. Leverage Conversation History
ChatBox saves your conversation history:
- The AI remembers the context of the current session
- You can keep asking based on earlier answers
- Great for in-depth discussion and iteration
4. Switch Models
Switch to the right model for different tasks:
- Writing tasks - use
gpt-4oorclaude-sonnet-4-5 - Coding tasks - use
claude-sonnet-4-5 - Quick Q&A - use
gpt-4o-mini - Multimodal - use
gemini-2.0-flash-exp
Features
With ChatBox + MaiToken, you can:
- 💬 Smooth conversation - real-time streaming for a smoother experience
- 🎯 Multi-session management - manage multiple independent conversations at once
- 💾 Local storage - conversations are stored locally to protect your privacy
- 📤 Export conversations - export conversation records in multiple formats
- 🎨 Clean interface - a clean, attractive UI
- 🔒 Open source and free - fully open source and free to use
- 🌍 Cross-platform - supports Windows, macOS, and Linux
- 🚀 Great performance - a lightweight app that runs smoothly
Support and Help
If you run into any issues while using MaiToken:
