# Cherry Studio Quick Start

## Purpose

This guide connects Cherry Studio to AnyRoute and verifies the connection.

## Quick configuration

You need three values:

- **Base URL:** `https://cc.anyroute.io`
- **API key:** a key created in AnyRoute
- **Model ID:** for example, `claude-opus-4-6`

## 1. Before you begin

Confirm that Cherry Studio is installed and that you have an AnyRoute account and API key.

### Create an API key

1. Sign in to the [AnyRoute API Keys page](https://cc.anyroute.io/keys).
2. Select **Create Key**.
3. Enter a name such as `cherry-studio`.
4. Select a group that matches the provider type and models you plan to use.
5. Keep the other settings at their defaults unless you need restrictions.

![Create an AnyRoute key](https://r2.bozhouai.com/dragoncode-cherry-studio/page2_img1.png)

![Copy the API key](https://api.apifox.com/api/v1/projects/8582374/resources/675913/image-preview)

The copied key must be complete and contain no added spaces.

## 2. Open Cherry Studio settings

Open Cherry Studio and select **Settings**.

![Cherry Studio settings](https://r2.bozhouai.com/dragoncode-cherry-studio/page4_img1.png)

## 3. Add AnyRoute as a provider

1. Under model services, select **Add**.

   ![Add a provider](https://r2.bozhouai.com/dragoncode-cherry-studio/page5_img1.png)

2. Enter a provider name such as `AnyRoute`.

   ![Enter the provider name](https://r2.bozhouai.com/dragoncode-cherry-studio/page5_img2.png)

3. Select **OpenAI** or **Anthropic** according to the model family and key group.

   ![Select a provider type](https://r2.bozhouai.com/dragoncode-cherry-studio/page6_img1.png)

4. Enter:

   - **API key:** your AnyRoute key
   - **API address:** `https://cc.anyroute.io`

   ![Enter the AnyRoute API settings](https://api.apifox.com/api/v1/projects/8582374/resources/675914/image-preview)

5. Add the exact model ID.

> Match the key group to the provider type: use a Codex/OpenAI-compatible group for OpenAI-type models and a Claude/Anthropic-compatible group for Anthropic-type models.

## Supported models

### OpenAI-compatible examples

- `gpt-5.4`
- `gpt-5.5`
- `gpt-5.6-sol`
- `gpt-5.6-terra`
- `gpt-5.6-luna`

### Anthropic-compatible examples

- `claude-opus-4-8`
- `claude-sonnet-5`
- `claude-fable-5`

For the full current list, see [Supported Models](https://en-guide.anyroute.io/9233271m0.md).

![Add a model](https://api.apifox.com/api/v1/projects/8582374/resources/675915/image-preview)

Select **Check** or **Test**. Continue when Cherry Studio reports a successful connection.

![Successful provider test](https://api.apifox.com/api/v1/projects/8582374/resources/675916/image-preview)

## 4. Verify with a chat

1. Create a new conversation.
2. Select the model you added.
3. Send a short test message such as `Hello`.

![Select the AnyRoute model](https://api.apifox.com/api/v1/projects/8582374/resources/675918/image-preview)

![Successful model response](https://api.apifox.com/api/v1/projects/8582374/resources/675917/image-preview)

## Troubleshooting

### HTTP 503

The selected model route may be unavailable or busy. Retry later or select another available model.

### HTTP 502

The upstream model connection failed. Wait briefly and retry, or select another model.

### `AI_RetryError`

Automatic retries were exhausted, usually after repeated upstream errors. Try another model or retry later.

### HTTP 401 or 403

Copy the key again and confirm that the provider type matches the key group.

### HTTP 400

Check that the API address is `https://cc.anyroute.io` and that the model ID exactly matches a supported model.

### The request keeps loading

Start a new conversation, restart Cherry Studio, check the network connection, or switch models and groups.

## One-minute checklist

| Setting | Expected value |
|---|---|
| Base URL | `https://cc.anyroute.io` |
| API key | Complete, active, and without spaces |
| Model ID | Exact supported model ID |
| Group and type | OpenAI-compatible group ↔ OpenAI; Claude-compatible group ↔ Anthropic |