Chatbox

A cross-platform desktop chat client. The simplest to configure, and a good fit for everyday conversation and quick model trials.

Download: chatboxai.app — Windows (x64), macOS (Intel / Apple Silicon) and Linux (x86_64 / ARM64) are all supported.

If macOS blocks the first launch, allow it under System Settings → Privacy & Security.

Before you start

You need an API Key with available quota. Copy model names from the model list — they are case-sensitive.

Configuration

The interface is essentially identical across all three platforms.

1. Click Settings in the bottom-left of the main window

2. Choose Model Provider in the left menu, then click Add

3. Fill in:

   Field	What to enter
   Provider name	TokenRoute (your choice)
   API mode	OpenAI API Compatible
   API Key	Your API Key
   API host	https://api.smartwan.com
   API path	Leave as default

4. Click Fetch models to pull in every available model

5. Close settings, click New conversation on the left, and pick a TokenRoute model in the bottom-right

Verify

Start a new conversation and send:

Hello, please reply: Chatbox setup successful

A normal reply means you're done.

Enabling image and tool capabilities

Text only by default

Chatbox enables text input only by default. Image understanding, function calling and similar capabilities must be switched on manually in the model settings.

Click the ⚙️ next to a model to open its detailed settings, then tick the input capabilities you need — images, tools, and so on.

You can also click the Check button next to the API Key to see which call types the current model actually supports.

Troubleshooting

Symptom	Cause
Model list comes back empty	Wrong API Key, or the account has no quota
404 error	API host was set to https://api.smartwan.com/v1 — remove the /v1
Model selectable but replies error out	That model may be outside this key's allowed set

More in the FAQ.