Building a Local AI Assistant with DeepSeek and Chatbox Using Ollama

If you want to quickly build a local client‑side AI assistant, replace OpenAI with DeepSeek to experience a domestic solution, and you are a beginner who prefers a simple setup, this guide will take you from zero to a working assistant in about ten minutes. Please like the article so we can publish more beginner‑friendly AI tutorials.

Preface

The title is taken from DeepSeek to illustrate its appealing quality.

1. Install DeepSeek

Installation of DeepSeek relies on Ollama.

Ollama is an open‑source framework designed for easy deployment and execution of large language models (LLMs) on local machines. It simplifies running LLMs inside Docker containers, allowing non‑experts to manage and operate complex models. Ollama currently supports macOS, Linux, and Windows; download the appropriate installer from the official site.

https://ollama.com/

After the installation finishes, search for the DeepSeek model in Ollama, copy the command shown in the screenshot, and run it in the console:

ollama run deepseek-r1:7b

Once the model is running, your computer becomes a small DeepSeek server and you can start asking it questions.

If you have multiple models on your machine, you can manage them with the list and run commands.

2. Install Chatbox

Chatbox AI is an AI client application and smart assistant that supports many advanced AI models and APIs, and runs on Windows, macOS, Android, iOS, Linux, and the web.

After downloading, launch the application and first set the language to Chinese.

Then click the model selection and choose the DeepSeek model you started earlier.

After saving, you can happily use the various features of Chatbox with the DeepSeek model.

Scan the QR code to join the technical discussion group