Step-by-Step Guide to Deploy DeepSeek AI Locally on macOS with Ollama and Chatbox AI

DeepSeek, developed by Hangzhou DeepSeek AI Technology Co., Ltd., is an advanced artificial‑intelligence product whose R1 model rivals ChatGPT in performance and is completely free.

Local deployment of DeepSeek is recommended to avoid server overload, improve stability, reduce latency, and protect sensitive data by keeping the model and its outputs on the user’s machine.

Installation steps for macOS:

1. Download Ollama from the official site https://ollama.com/download (or use the provided backup link if needed).

2. Move the downloaded application into the /Applications folder.

3. Open the app and click Next , then Install , and finally Finish .

4. Open Terminal (found in Applications → Utilities) and run ollama to verify the installation.

5. In the Ollama UI, go to Models and locate the deepseek-r1 model.

6. Copy the provided download command for the model and paste it into the terminal; note that the full R1 model requires about 404 GB of storage, so a smaller 1.5 B or 7 B version is recommended for most Macs.

7. Wait for the model to download; once completed you can start asking questions directly in the terminal prompt.

Visual integration with Chatbox AI:

1. Download the Chatbox AI visual tool from https://chatboxai.app/zh and move it to /Applications .

2. Launch Chatbox AI, choose “Use your own API Key or local model”, select the Ollama API , and set the model to deepseek-r1:1.5b .

3. After configuration, the visual interface allows local conversations with DeepSeek, providing fast, private responses.