# Cherry Studio API key: use Claude and GPT with 5abeer

> Connect Cherry Studio to 5abeer with a custom provider, sync your models, and set the Quick Model and context options that decide what each chat costs.

Updated: 2026-11-03

To use Cherry Studio with a 5abeer API key, open Settings, then Model Provider, and add a custom provider. Paste your key in API Key and enter `https://api.5abeer.ai` in the Anthropic field for a Claude key, or the OpenAI field for a GPT key. Then sync and add the models, switch the provider on and pick a model.

## What do you need before you start?

- **Cherry Studio**, the desktop app, on Windows, macOS or Linux. Download it from the [official download page](https://cherryai.com.cn/download/v2) for your system and chip. On 2026-11-03, the latest release was 2.1.10, published on 2026-10-31.
- **A 5abeer API key** for the provider of the models you want. A key works with one provider's models only ([Create a key](https://5abeer.ai/en/docs/start/keys.md#create)). New to keys? See [How to get an API key](https://5abeer.ai/en/guides/ai-api-key-guide).
- **Some balance** on your account ([Payment methods](https://5abeer.ai/en/docs/start/payment.md)).

The names below are from Cherry Studio's English interface; version 2.1.10 has no Arabic one.

A 5abeer plan won't work here. Plans are for the supported apps, and Cherry Studio is not one of them, so it runs on an API key paid from your balance.

## Which endpoint should you choose?

A custom provider asks which API format, or endpoint, to use. The window shows **OpenAI** and **Anthropic**, and **More options** adds **OpenAI Responses** and **Gemini**. Pick the one that matches your key:

- **Anthropic**, for a Claude key. 5abeer is compatible with the Anthropic Messages API.
- **OpenAI**, for a GPT key. 5abeer is OpenAI-compatible.
- **Gemini**, under **More options**, for a Gemini key.

Enter the same root address in any of them: `https://api.5abeer.ai`. Cherry Studio adds `/v1` for OpenAI and Anthropic and `/v1beta` for Gemini, and shows the final **Request path** under the field before you save. These are our base URLs:

| Format | Base URL | Key header |
| --- | --- | --- |
| OpenAI | `https://api.5abeer.ai/v1` | `Authorization: Bearer <KEY>` |
| Anthropic | `https://api.5abeer.ai` | `x-api-key: <KEY>` |
| Gemini | `https://api.5abeer.ai/v1beta` | `x-goog-api-key: <KEY>` |

## How do you set up Cherry Studio step by step?

### 1. Create a 5abeer API key

In your dashboard, open **API → Keys**, press **New key** and pick Claude. Copy the key when it appears.

### 2. Add a custom provider

In Cherry Studio, open **Settings** with the gear icon, then **Model Provider**, and press **Add Provider**. The **Add Custom Provider** window opens.

### 3. Fill in the fields

- **Provider Name**: a name you will recognise, such as `5abeer Claude`.
- **API Key**: your 5abeer key.
- **Endpoint settings**: enter `https://api.5abeer.ai` in **Anthropic** and leave **OpenAI** empty. The **Request path** should end in `/v1/messages`.

Press **Add**.

### 4. Sync the models

On the provider's page, press **Sync models**. Cherry Studio asks 5abeer for the models your key can use and lists them under **Model management**. Add the ones you want with **+**: only added models appear in a chat's model picker. If one is missing, press the **+** next to **Sync models** (**Add model manually**) and type its **Model ID** exactly as on the [models list](https://5abeer.ai/en/models).

### 5. Check the connection

Next to **API Key**, press **Model Check**, pick one model under **Select Model** and press **Start**. Cherry Studio sends a short real request with your key. If it fails, look in the provider's list for a mistyped model or one your key doesn't support.

### 6. Switch the provider on

Turn on the switch at the top right of the provider's page. Its models appear in chats only while it is on.

### 7. Send a test message

Open a new chat, choose [`claude-sonnet-5.5`](https://5abeer.ai/en/models/claude-sonnet-5.5) and send a short message. If a reply comes back, the setup works.

## Which model should you use in Cherry Studio?

Start with [`claude-sonnet-5.5`](https://5abeer.ai/en/models/claude-sonnet-5.5) for everyday chat, writing and coding questions. Move to [`claude-opus-5.5`](https://5abeer.ai/en/models/claude-opus-5.5) for long, hard tasks where the extra cost pays off.

Set `claude-sonnet-5.5` as the **Default Assistant Model** in **Settings → Default Model**, and give only your hard-task assistant its own model, `claude-opus-5.5`. An assistant's own model overrides the default. These are the current Claude models:

| Model | Id | Input (ours) | Output (ours) | Context |
| --- | --- | --- | --- | --- |
| Claude Sonnet 5.5 | `claude-sonnet-5.5` | $0.60 | $3 | 1M |
| Claude Opus 5.5 | `claude-opus-5.5` | $1.20 | $6 | 1M |
| Claude Fable 5.1 | `claude-fable-5.1` | $3 | $15 | — |

The official price next to ours, with the date it was read:

Claude Sonnet 5.5 (`claude-sonnet-5.5`) per 1M tokens: input $0.60 (official $2), output $3 (official $10). Official price checked October 19, 2026.

GPT and Gemini models need a key for their own provider. Their current chat models:

| Model | Id | Input (ours) | Output (ours) | Context |
| --- | --- | --- | --- | --- |
| GPT-6.1 Sol | `gpt-6.1-sol` | $0.60 | $3 | 922K |
| GPT-6 Astra | `gpt-6-astra` | $3 | $15 | — |
| GPT-6 Sol | `gpt-6-sol` | $0.60 | $3 | 922K |

| Model | Id | Input (ours) | Output (ours) | Context |
| --- | --- | --- | --- | --- |
| Gemini 4 Argon | `gemini-4-argon` | $0.60 | $3 | — |
| Gemini 2.5 Pro | `gemini-2.5-pro` | $0.38 | $3 | 1M |

Our prices next to the official ones: [Claude API](https://5abeer.ai/en/api/claude) and [OpenAI API](https://5abeer.ai/en/api/openai).

## Which Cherry Studio settings decide what you pay?

Each request is paid from your balance by the tokens it uses, at the model's 5abeer price ([Pricing and balance](https://5abeer.ai/en/docs/start/pricing.md)). Through the API, you use every model priced per token for less than its official price.

Three settings decide how big your chat requests are and which model runs the requests Cherry Studio sends on its own. We checked them against version 2.1.10 on 2026-11-03:

- **Quick Model**, in **Settings → Default Model**. Cherry Studio uses it for small tasks such as naming conversations and extracting search keywords. While it is empty, it follows the Default Assistant Model, and a default you pick while it is empty is copied into it. So look at what it shows: if it is `claude-opus-5.5`, that model writes every conversation title. Set it to `claude-sonnet-5.5`. To stop naming requests, press the gear next to Quick Model and turn off **Conversation Auto Naming**. Translate Model follows the same rule.
- **Recent messages kept**, in **Settings → General → Context Management**. It is empty by default, so each new message carries the whole chat with it. A number such as 20 sends only the latest 20 messages, though too small a number makes the model forget earlier instructions. Each assistant can set its own value.
- **Auto compress**, on the same page, is on by default. When a chat nears the model's context limit, Cherry Studio summarises the older turns with the chat's own model. Pick `claude-sonnet-5.5` as the **Compression model**, so long chats on `claude-opus-5.5` are summarised at the lower price.

Two more habits:

- **Check one model, not all.** Each check is a real, paid request, and Cherry Studio warns that **Check all models** can cost a lot in a short time.
- **Give Cherry Studio its own key**, with a spend limit, an expiry date and an IP allow-list ([Key controls](https://5abeer.ai/en/docs/start/keys.md#controls)).

## Can you use Claude and GPT in the same app?

Yes, with one Cherry Studio provider per key. A 5abeer key works with one provider's models, and a GPT model called with a Claude key returns [`MODEL_PLATFORM_NOT_ALLOWED`](https://5abeer.ai/en/docs/help/errors.md#MODEL_PLATFORM_NOT_ALLOWED). To add GPT, create a GPT key, then add a second custom provider, such as `5abeer GPT`, with `https://api.5abeer.ai` in **OpenAI** and the GPT key in **API Key**. Sync, add, check and switch it on as above.

Both providers draw on the same balance.

Don't put keys for two providers in one Cherry Studio provider. Cherry Studio can hold several keys per provider and uses them in turn, so some requests would go out with the wrong key.

## Common errors and fixes

- [`INVALID_API_KEY`](https://5abeer.ai/en/docs/help/errors.md#INVALID_API_KEY) or a 401: the key has a typo or an extra space, or was revoked. Paste it again on the provider's page.
- [`MODEL_PLATFORM_NOT_ALLOWED`](https://5abeer.ai/en/docs/help/errors.md#MODEL_PLATFORM_NOT_ALLOWED): the model belongs to another provider than the key. Pick it under the Cherry Studio provider that holds that model's key.
- **Not found, or a wrong path:** the address holds a full path such as `/v1/chat/completions`, or ends in `#`, which stops Cherry Studio from adding the version. Enter only `https://api.5abeer.ai` and check the **Request path**.
- **No 5abeer models in a chat:** the provider is off, or you synced without adding models.
- **Model not found:** the Model ID you typed doesn't match ours. Copy it from the [models list](https://5abeer.ai/en/models).
- [`INSUFFICIENT_BALANCE`](https://5abeer.ai/en/docs/help/errors.md#INSUFFICIENT_BALANCE): top up, then retry.

Every code is explained in [Errors and fixes](https://5abeer.ai/en/docs/help/errors.md).

## Questions

### Should the address end in /v1?

It doesn't need to. Cherry Studio adds `/v1` itself, and if you enter `https://api.5abeer.ai/v1`, it sees the version and doesn't add it twice.

### Why does Sync models show only Claude models?

Because the list comes from your key, and a key works with one provider's models. To see GPT models, add a second provider with a GPT key.

### Why does my usage show requests I didn't type?

Those are Cherry Studio's own small tasks, such as naming a conversation or extracting search keywords. They run on the Quick Model and are paid like any request. Summaries of long chats also run in the background, on the Compression model. Point Quick Model at a lighter model, or turn off Conversation Auto Naming, as in [the cost settings](https://5abeer.ai/en/guides/cherry-studio-api-key#costs).

### Can I use my key in Cherry Studio's agent?

Yes. In version 2.1.10, an agent's **Runtime mode** is **Claude Agent**, which uses the Anthropic format of your Claude provider, or **Pi**, which also works with OpenAI-format providers such as one holding a GPT key. The agent's requests are paid from your balance, so set a spend limit on that key.

> Running a chat app on your own server for a team instead? See [Open WebUI](https://5abeer.ai/en/guides/open-webui-api-key) or [LibreChat](https://5abeer.ai/en/guides/librechat-api-key). What the base URL field means in any app is in [What is a base URL?](https://5abeer.ai/en/guides/what-is-base-url)

## Sources

Checked on 2026-11-03:

- Cherry Studio docs, in Chinese: [Model provider settings](https://docs.cherryai.com.cn/pre-basic/providers/providers), [Custom provider](https://docs.cherryai.com.cn/pre-basic/providers/zi-ding-yi-fu-wu-shang), [Anthropic](https://docs.cherryai.com.cn/pre-basic/providers/anthropic), [Default models](https://docs.cherryai.com.cn/pre-basic/settings/default-models), [Long chats and context](https://docs.cherryai.com.cn/advanced-basic/chat/context-queue) and [Installation](https://docs.cherryai.com.cn/cherry-studio/installation)
- Cherry Studio on GitHub: [release v2.1.10](https://github.com/CherryHQ/cherry-studio/releases/tag/v2.1.10), and the source at that tag for the [interface languages](https://github.com/CherryHQ/cherry-studio/blob/v2.1.10/src/renderer/i18n/languages.ts), [English labels](https://github.com/CherryHQ/cherry-studio/blob/v2.1.10/src/renderer/i18n/locales/en-us.json), [Model Check](https://github.com/CherryHQ/cherry-studio/blob/v2.1.10/src/renderer/pages/settings/ProviderSettings/ModelList/ModelCheckDialog.tsx), [Quick Model fallback](https://github.com/CherryHQ/cherry-studio/blob/v2.1.10/src/renderer/hooks/useModel.ts) and [agent run modes](https://github.com/CherryHQ/cherry-studio/blob/v2.1.10/src/renderer/components/AgentRuntimeOption.tsx)
- 5abeer docs: [API keys](https://5abeer.ai/en/docs/start/keys.md), [Base URLs](https://5abeer.ai/en/docs/start/base-urls.md), [List models](https://5abeer.ai/en/docs/api/models.md) and [Errors and fixes](https://5abeer.ai/en/docs/help/errors.md)

> 5abeer is an independent service, not affiliated with Cherry Studio, OpenAI, Anthropic or Google.