# LobeHub Setup Guide for OfoxAI

Source: https://ofox.ai/docs/integrations/lobehub

[LobeHub](https://lobehub.com) is a powerful open-source AI desktop client (macOS, Windows, Linux, Web) that supports multiple AI service providers.

OfoxAI supports four protocols in LobeHub:

| Protocol | Provider | API Proxy URL |
|----------|----------|--------------|
| **OpenAI Chat** (recommended) | OpenAI | `https://api.ofox.ai/v1` |
| **OpenAI Response** | OpenAI | `https://api.ofox.ai/v1` |
| **Claude** | Anthropic | `https://api.ofox.ai/anthropic` |
| **Gemini** | Google | `https://api.ofox.ai/gemini` |

## Prerequisites

- An OfoxAI account with an API Key ([Get one here](https://app.ofox.ai))
- LobeHub installed ([Download](https://lobehub.com/downloads))

## Setup Steps

### Step 1: Open Settings

Launch LobeHub and click the **Settings** icon in the bottom left.

![Open LobeHub Settings](https://ofox.ai/screenshots/lobehub/01-home.png)

### Step 2: Go to AI Service Providers

In the settings sidebar, click **AI Service Providers**, then select the provider you want to configure.

![Go to AI Service Providers](https://ofox.ai/screenshots/lobehub/02-ai-provider.png)

### Step 3: Fill in the Configuration

Select the provider and fill in the details. Example using **OpenAI Chat**:

| Field | Value |
|-------|-------|
| **API Key** | Your OfoxAI API Key |
| **API Proxy URL** | `https://api.ofox.ai/v1` |
| **Use Responses API** | Off (OpenAI Chat) / On (OpenAI Response) |

API Proxy URL by protocol:

| Protocol | Provider | API Proxy URL |
|----------|----------|--------------|
| **OpenAI Chat / Response** | OpenAI | `https://api.ofox.ai/v1` |
| **Claude** | Anthropic | `https://api.ofox.ai/anthropic` |
| **Gemini** | Google | `https://api.ofox.ai/gemini` |

After filling in, select a model in the connectivity check field and click the check button. "**Check passed**" means the configuration is successful.

![Fill in configuration and check connection](https://ofox.ai/screenshots/lobehub/03-openai-config.png)

**OpenAI Chat vs OpenAI Response:** When "Use Responses API" is off, requests go to `/v1/chat/completions` (Chat). When on, requests go to `/v1/responses` (Response, GPT series only).

## Start Using

After setup, go back to the main screen, click the model icon on the left of the input box to select a model, and start chatting.

![Start chatting](https://ofox.ai/screenshots/lobehub/04-start-chat.png)

## Troubleshooting

**Q: Connectivity check fails**

1. Make sure the API Proxy URL is correct (see table above)
2. Make sure your API Key is copied in full from the [OfoxAI Console](https://app.ofox.ai)
3. Check your network connection

**Q: What's the difference between OpenAI Chat and OpenAI Response?**

Both use the same API Proxy URL `https://api.ofox.ai/v1`, but differ in request format:

- **OpenAI Chat** (Responses API off): uses `/v1/chat/completions`, supports all models, best compatibility
- **OpenAI Response** (Responses API on): uses `/v1/responses`, GPT series only, but supports new features like `instructions` caching — better for high-frequency use cases

**Q: The model I want isn't in the list**

Click "Fetch model list" to auto-fetch, or click **+** to add a custom model. Use the `provider/model-name` format for the model ID.

See all models at [Models](https://ofox.ai/models).