Using Tokenhot in Cherry Studio#
Cherry Studio is an AI desktop client supporting Windows, macOS, and Linux, featuring built-in multi-model conversations, knowledge bases, and more.By configuring Tokenhot as a provider, you can access mainstream large models such as GPT, Claude, Gemini, and DeepSeek.Supported Access Protocols#
Tokenhot is compatible with the following four protocols:OpenAI Chat — Path /v1/chat/completions, for GPT series, e.g., openai/gpt-4.1.OpenAI Response — Path /v1/responses, also for GPT series, e.g., openai/gpt-5.4-mini.Claude — Path /anthropic/v1/messages, for Claude series, e.g., anthropic/claude-sonnet-4.6.Gemini — Path /gemini/v1beta/models, for Gemini series, e.g., google/gemini-3.1-pro-preview.Prerequisites#
Configuration Steps#
Step 1: Open Settings#
Launch Cherry Studio and click the Settings icon in the bottom-left corner.Step 2: Add a Provider#
Go to Model Services, click + Add at the bottom, enter a name, and select the corresponding type:OpenAI Chat / OpenAI Response → Type: OpenAI
For OpenAI Response, you need to additionally enable Response mode on the configuration page.
The API addresses for each protocol are as follows:| Protocol | API Address |
|---|
| OpenAI Chat / Response | https://api.tokenhot.ai/v1 |
| Claude | https://api.tokenhot.ai |
| Gemini | https://api.tokenhot.ai |
Enter your Tokenhot API Key uniformly as the API key.Ensure the toggle switch in the top-right corner is ON, otherwise the provider will not appear in the model selector.
Step 4: Add Models#
Click the Manage button to fetch the model list, then click the + icon next to a model to add it to the enabled list.Step 5: Verify#
Click Test, select a model to test — if it passes, the configuration is complete.Start Chatting#
Return to the main interface, click the model selector at the top, and choose a model under the corresponding provider to start using it.Image Generation#
Cherry Studio supports invoking Tokenhot's image generation capabilities via the Gemini protocol.Configuration Steps#
1.
Add a Gemini-type provider with the API address https://api.tokenhot.ai
2.
Add the desired image generation model
4. Right-click an assistant → Edit Assistant, switch the default model to the image generation model, and add the following custom parameters:| Parameter | Type | Value |
|---|
responseModalities | JSON | ["IMAGE"] |
imageConfig | JSON | {"aspectRatio": "1:1", "imageSize": "1K"} |
For 2K resolution, change imageSize to "2K".4.
Save, then enter a description in the chat box to generate the image directly.
FAQ#
Empty model list: Check that the API address has no trailing slash and that the API Key is pasted completely.Connection failed: Verify the API Key and address are correct, and check whether your local network is working properly.Provider not showing: Confirm the toggle switch on the configuration page is turned on.Empty model list: Manually add the desired models.Modified at 2026-05-29 02:58:02
