Use FuturMix with Chatbox

Configure the Chatbox desktop AI client to use FuturMix as your API provider. Takes about 3 minutes.

Updated June 4, 2026

Prerequisites

  1. FuturMix accountSign up and top up with at least $10.
  2. API Key — Log in, go to Console → API Keys, and create a key (starts with sk-).
  3. Chatbox app — Download the desktop version from chatboxai.app. Windows, macOS, and Linux are all supported.

Setup Steps

1

Open Settings

Launch Chatbox and click the gear icon in the lower-left corner of the sidebar.

2

Add a Provider

Navigate to Settings → Model Provider and click "+ Add".

3

Enter Configuration

NameFuturMix
API ModeOpenAI API Compatible
API Hosthttps://futurmix.ai/v1
API Path(leave empty)
API Keysk-your-key-here
4

Verify Connection

Click the "Check" button next to the API Key field. A success message confirms your configuration is correct.

If your account balance is $0, the Check button will return "Connection Failed" even if credentials are correct. Top up first, then retry.
5

Add Models

Click "Fetch" to auto-load the available model list, or click "+ New" and type a model ID manually.

Recommended models by use case:

Use CaseModelInput / Output (per 1M tokens)Discount
Everyday chatgpt-5.4-nano$0.14 / $0.8830% off
Code & analysisgpt-5.4$1.75 / $10.5030% off
Deep reasoningclaude-opus-4-6$4.50 / $22.5010% off
Fast responsesclaude-haiku-4-5-20251001$0.90 / $4.5010% off
Best valuegemini-2.5-flash$0.24 / $2.0020% off
When adding a model, you can check capability flags: vision (image input), tool_use (function calling / MCP tools), and reasoning (thinking indicator). GPT-5.4, Claude Sonnet/Opus, and Gemini models support vision.
6

Start Chatting

Return to the main screen, select a FuturMix model from the model picker, and send a message.

FAQ

What should I enter for API Host?

Enter https://futurmix.ai/v1 and leave API Path empty. Chatbox constructs the full endpoint as https://futurmix.ai/v1/chat/completions.

What if Fetch doesn't load models?

Click "+ New" and type the model ID directly (e.g., gpt-5.4, claude-opus-4-6). Chatbox accepts any model ID you enter manually.

Does it support images and vision?

Yes. Check the vision capability flag when adding a model. You can then drag and drop images into the chat. Supported models include GPT-5.4, Claude Sonnet/Opus series, and Gemini series.

Does it support tool use and function calling?

Yes. Check the tool_use capability flag. This enables Chatbox's built-in web search and MCP tool integrations.

Does it work on iOS?

The desktop app (Windows, macOS, Linux) is the most reliable. iOS may encounter CORS issues with custom API endpoints due to a known Chatbox limitation. Android works normally.

How do I check my balance?

Log in at futurmix.ai and go to Console → Dashboard.

Can I use multiple models at once?

Yes. Add as many models as you want under your FuturMix provider. Switch between them in the model picker at any time.

How is this different from OpenRouter?

Both are OpenAI-compatible multi-model gateways. The setup is identical — change the API Host and key to switch. FuturMix offers 10–30% discounts off official model pricing with no platform fee. OpenRouter charges a 5.5% fee on top of model pricing.

Troubleshooting

SymptomCauseFix
Check returns "Connection Failed"$0 balance or invalid keyTop up (≥$10) and regenerate your API key in Console
401 UnauthorizedWrong or expired API keyCreate a new key in Console → API Keys
404 Not FoundAPI Host format wrongUse https://futurmix.ai/v1 (include /v1)
Empty responseModel ID typoUse Fetch to load the correct model list
Images not workingVision flag not checkedEdit the model and enable the vision capability

Ready to get started?

Create an account, top up from $10, and start using 20+ AI models in Chatbox.

No subscription required. Pay only for what you use.