Are you looking to leverage the full potential of DeepSeek, a powerful AI large language model (LLM), but find the official website a bit sluggish? This guide walks you through configuring DeepSeek with alternative platforms like Cherry Studio and Chatbox, bypassing those frustrating lags. We will focus on API configuration, the most cost-effective solution for individual users.
This article focuses on using the DeepSeek API, offering a more stable and customizable experience. We'll explore the advantages of API usage, step-by-step configuration for Cherry Studio and Chatbox, and a detailed FAQ to address common questions.
While the DeepSeek web interface (DeepSeek - R1) is readily accessible, it can be prone to performance issues due to high traffic. Configuring DeepSeek through an API offers several advantages:
Cherry Studio is a desktop application that provides a clean and efficient interface for interacting with AI models.
获取步骤 | API 域名 | 模型 - 聊天 | 模型 - 推理 | 设置 - 推理 | |
---|---|---|---|---|---|
硅基流动 | 图文 | https://api.siliconflow.cn | deepseek-ai/DeepSeek-V3 | deepseek-ai/DeepSeek-R1 | |
阿里云百炼 | 图文 | https://dashscope.aliyuncs.com/compatible-mode/v1/ | deepseek-v3 | deepseek-r1 | |
百度智能云/百度云千帆 | 图文 | https://qianfan.baidubce.com/v2/ | deepseek-v3 | deepseek-r1 | |
字节火山引擎 | 图文 | https://ark.cn-beijing.volces.com/api/v3/ | 访问推理点获取 | 访问推理点获取 | 访问推理点获取 |
Chatbox is a versatile application available on desktop and mobile platforms, making it a great option for accessing DeepSeek on the go.
Click on “Add Custom Provider” via:
获取步骤 | API 域名 | 模型 - 聊天 | 模型 - 推理 | 设置 - 推理 | |
---|---|---|---|---|---|
硅基流动 | 图文 | https://api.siliconflow.cn/v1 | deepseek-ai/DeepSeek-V3 | deepseek-ai/DeepSeek-R1 | |
阿里云百炼 | 图文 | https://dashscope.aliyuncs.com/compatible-mode/v1 | deepseek-v3 | deepseek-r1 | |
百度智能云 | 图文 | https://qianfan.baidubce.com/v2 | deepseek-v3 | deepseek-r1 | |
字节火山引擎 | 图文 | https://ark.cn-beijing.volces.com/api/v3 | 访问推理点获取 | 访问推理点获取 | 访问推理点获取 |
Q1: How do I add more models from a custom provider (e.g., Silicon Flow)? (Cherry Studio)
In Cherry Studio, select Silicon Flow, click “Add” at the bottom, input the Model ID, and save. The model name and group name would be filled automatically upon entering the ID.
Q2: How do I change parameters like temperature? (Cherry Studio)
In Cherry Studio, click "Settings" on the left side of the chat interface to adjust the model temperature and context window size.
Q3: How do I back up my data? (Cherry Studio)
In Cherry Studio, go to Settings -> Data Settings -> Backup
Q4: How do I change the default model? (Cherry Studio)
In Cherry Studio, go to Settings -> Default Model -> Default Assistant Model
Q5: How do I modify the default assistant template? (Cherry Studio)
In Cherry Studio, go to Settings -> Default Model -> Default Assistant Model, click the settings icon on the right.
Q1: How to Add Models to Custom Provider? (Chatbox) In Chatbox, go to settings and enter a new model name, select a model, click the “+” sign and Save.
Q2: How do I switch platforms? (Chatbox)
In Chatbox, click Settings, and select the desired platform from the dropdown list, and Save.
Q3: How to I share Chat Logs? (Chatbox)
In Chatbox, go to Settings -> Other -> Data backup -> Check Items for export -> Export Selected Items
Q4: How should the temperature be set? (Chatbox) The default temperature is 1.0. The best setting depends on the use case:
Scenario | Temperature |
---|---|
Code Generation/Math | 0.0 |
Data Extraction Analysis | 1.0 |
General Conversations | 1.3 |
Translation | 1.3 |
Creative Writing/Poetry | 1.5 |
Important Note: The temperature
, top_p
, presence_penalty
, frequency_penalty
, logprobs
, top_logprobs
input parameters will not take effect when using the deepseek-reasoner
model.
By configuring DeepSeek with Cherry Studio or Chatbox using the API, you can enjoy a significantly improved user experience. You'll bypass the potential performance bottlenecks of the official website, gain greater control over customization, and leverage the generous token offerings from various platforms. So take control and unleash DeepSeek's full potential today!
Further Reading: