> ## Documentation Index
> Fetch the complete documentation index at: https://docs.shuyou.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# CodeBuddy / WorkBuddy

> Connect CodeBuddy and WorkBuddy to ShuYou via models.json and the OpenAI-compatible API.

# CodeBuddy / WorkBuddy

[CodeBuddy](https://www.codebuddy.cn/) and WorkBuddy are AI coding tools from Tencent Cloud. Both support custom model providers through a `models.json` configuration file. Connect them to **ShuYou** to use ShuYou's model catalog with a single API key.

<Info title="Applies to both tools">
  CodeBuddy and WorkBuddy use the same configuration format. The steps below apply to both.
</Info>

## Prerequisites

1. Install [CodeBuddy](https://www.codebuddy.cn/setup/) or WorkBuddy.
2. Create a ShuYou API key in the [console](https://shuyou.ai/signin).

ShuYou OpenAI-compatible chat completions endpoint: **`https://api.shuyou.ai/v1/chat/completions`**

<Note>
  Currently, CodeBuddy / WorkBuddy integration with ShuYou uses the **OpenAI SDK format** (Chat Completions). Use model slugs from the [ShuYou model list](https://shuyou.ai/models) that are labeled **OpenAI API Compatible**.
</Note>

## Configuration

### Step 1: Open the configuration file

**CodeBuddy:** `~/.codebuddy/models.json`

**WorkBuddy:** `~/.workbuddy/models.json`

On Windows:

* CodeBuddy: `C:\Users\<username>\.codebuddy\models.json`
* WorkBuddy: `C:\Users\<username>\.workbuddy\models.json`

### Step 2: Add ShuYou model entries

Edit `models.json` and add entries like the following:

```json theme={null}
{
  "models": [
    {
      "id": "ShuYou/auto",
      "name": "ShuYou Auto (Smart Routing)",
      "vendor": "ShuYou",
      "apiKey": "YOUR_SHUYOU_API_KEY",
      "url": "https://api.shuyou.ai/v1/chat/completions",
      "supportsToolCall": true,
      "supportsImages": true
    },
    {
      "id": "openai/gpt-5",
      "name": "ShuYou GPT-5",
      "vendor": "OpenAI",
      "apiKey": "YOUR_SHUYOU_API_KEY",
      "url": "https://api.shuyou.ai/v1/chat/completions",
      "supportsToolCall": true,
      "supportsImages": true
    },
    {
      "id": "anthropic/claude-sonnet-4.5",
      "name": "ShuYou Claude Sonnet 4.5",
      "vendor": "Anthropic",
      "apiKey": "YOUR_SHUYOU_API_KEY",
      "url": "https://api.shuyou.ai/v1/chat/completions",
      "supportsToolCall": true,
      "supportsImages": true
    }
  ]
}
```

<Warning title="Replace the API Key">
  Replace `YOUR_SHUYOU_API_KEY` with your real ShuYou API key from the [console](https://shuyou.ai/signin).
</Warning>

| Field                                 | Description                                               |
| ------------------------------------- | --------------------------------------------------------- |
| `id`                                  | Model slug from [ShuYou models](https://shuyou.ai/models) |
| `name`                                | Display name in the model dropdown                        |
| `url`                                 | Must be `https://api.shuyou.ai/v1/chat/completions`       |
| `supportsToolCall` / `supportsImages` | Set based on the model's capabilities                     |

### More available models

Add more objects to the `models` array using the same format. Example slugs (verify availability on ShuYou):

**GPT series:** `openai/gpt-5`, `openai/gpt-4o`

**Claude series:** `anthropic/claude-sonnet-4.5`, `anthropic/claude-haiku-4.5`

**Gemini series:** `google/gemini-2.5-pro`, `google/gemini-2.5-flash`

**Other:** `deepseek/deepseek-chat`, `qwen/qwen3-coder-plus`

See the full list on [shuyou.ai/models](https://shuyou.ai/models).

### Step 3: Save and reload

After saving `models.json`, the tool detects changes and reloads automatically (about 1 second debounce). Select a ShuYou model from the model dropdown to start.

## Using ShuYou Auto smart routing

### What is ShuYou Auto?

`ShuYou/auto` is ShuYou's intelligent routing model. The platform selects a suitable model based on request content and your routing configuration.

### Benefits

* **Smart matching** — Analyzes the task and picks an appropriate model
* **Cost optimization** — Balances quality and price
* **Load balancing** — Distributes traffic across the model pool
* **Transparent** — The response includes the actual model used

### Usage

Add an entry with `"id": "ShuYou/auto"` as shown above, then select **ShuYou Auto (Smart Routing)** in the model dropdown.

For advanced routing options, see [Model Routing](/en/guide/advanced/model-routing).

## Limit the model dropdown

To show only specific models, use `availableModels`:

```json theme={null}
{
  "models": [
    // ... model entries
  ],
  "availableModels": [
    "ShuYou/auto",
    "openai/gpt-5",
    "anthropic/claude-sonnet-4.5"
  ]
}
```

## Project-level configuration

In addition to user-level config, you can add project-scoped files:

* **CodeBuddy:** `<project-root>/.codebuddy/models.json`
* **WorkBuddy:** `<project-root>/.workbuddy/models.json`

<Info title="Priority">
  Project-level configuration overrides user-level settings. Use user-level config for global models and project-level config for repo-specific models.
</Info>

## FAQ

### Configuration not applied?

1. Validate JSON syntax (use a JSON linter).
2. Confirm the API key and model slugs are correct.
3. Restart the application if changes are not picked up.

### Which models are supported?

Any model on ShuYou that supports the OpenAI Chat Completions protocol. Filter **OpenAI API Compatible** on the [model list](https://shuyou.ai/models).

### Is the API key secure?

The key is stored locally in `models.json`. Restrict file permissions so other users on the machine cannot read it.

## Related links

* [ShuYou console](https://shuyou.ai/signin)
* [ShuYou model list](https://shuyou.ai/models)
* [Create Chat Completion](/en/api-reference/language-series/openai/create-chat-completion)
* [CodeBuddy documentation](https://www.codebuddy.cn/docs)