Complete Tutorial on Integrating Cherry Studio with Token Circle

Cherry Studio is an open-source AI client tool that supports connecting to various OpenAI-compatible API services. This article will detail how to connect Token Circle to Cherry Studio, completing configuration in 5 minutes.

---

Introduction to Cherry Studio

Cherry Studio is a cross-platform desktop AI client that supports Windows, macOS, and Linux. Its core features include:

• Multi-model switching: Quickly switch between different AI models within the same interface
• OpenAI compatible: Supports all services compatible with the OpenAI API format
• Session management: Supports multi-turn conversations, conversation history, export, and more
• Local-first: Data is stored locally, ensuring privacy
• Free and open-source: Completely free to use, with transparent and open code

---

Step 1: Download and Install Cherry Studio

Visit the official Cherry Studio website and download the version suitable for your operating system:

• Windows: Download the .exe installer and double-click to install
• macOS: Download the .dmg file and drag it into the Applications folder
• Linux: Download the .AppImage or .deb package

Once installed, launch Cherry Studio and you will see a clean chat interface.

---

Step 2: Configure Token Circle API

2.1 Add Provider

In the left menu of Cherry Studio, find "Settings", then click "Providers" or "API Providers". Click the "Add" button and select the "OpenAI Compatible" type.

2.2 Fill in Configuration Information

In the pop-up configuration window, fill in the following information:

Configuration Item	Content
Name	Token Circle or any name
API Base URL	https://www.ciyuano.com/v1
API Key	The API Key obtained from the Token Circle user center
Model	Add manually or click "Get model list"

2.3 Save and Test

After filling in, click "Save". Then return to the chat interface, select the provider and model you just added from the top model selector, send a test message, and confirm that you receive a reply normally.

---

Step 3: Choose the Right Model

Token Circle currently supports the following mainstream models. You can choose according to your needs:

Model	Features	Use Cases	Price
DeepSeek V4	Strong general ability, balanced Chinese and English	Daily conversation, translation, writing	0.5/1 yuan
MiMo v2.5 PRO	Strong reasoning ability, high code quality	Programming, mathematics, Agent	Free
GLM-5	Excellent Chinese understanding, rich knowledge	Chinese writing, knowledge Q&A	0.8/1.5 yuan
Qwen Plus	Good long context support	Long document analysis, summarization	0.4/0.8 yuan

---

Advanced Configuration

Add Multiple Models

You can add multiple models for the same provider. In the provider configuration, click the "Add" button next to the model list, and enter the model ID (e.g., deepseek-v4, mimo-v2.5-pro, glm-5).

Set System Prompt

In Cherry Studio's chat settings, you can configure different system prompts for different conversations. For example:

• Programming Assistant: "You are a professional programming assistant, proficient in Python, JavaScript, and other languages."
• Translation Assistant: "You are a professional translator, proficient in Chinese-English translation, maintaining the original style."
• Writing Assistant: "You are a writing assistant, helping users polish articles and generate content."

Adjust Parameters

In the chat settings, you can adjust the following parameters:

• Temperature: Controls the randomness of responses (0.1 for more precise, 0.9 for more creative)
• Max Tokens: Controls the maximum output length
• Top P: Controls the sampling range

---

FAQ

---

Summary

Through the steps above, you have successfully connected Token Circle to Cherry Studio. The entire process takes only 5 minutes:

1. Download and install Cherry Studio
2. Add an OpenAI-compatible provider, fill in the Token Circle API address and key
3. Choose the right model and start chatting

Cherry Studio's interface is clean and intuitive. Combined with Token Circle's multi-model support, you can easily switch between different models at any time to find the one best suited for your current task.