Deploying DeepSeek Locally on Windows to Another Drive

I’ve been using LM Studio to run DeepSeek locally, but the model files are too large and my C: drive is running out of space. It’s become unbearable. After some research, I found a solution using Ollama to download the model to another drive. Finally, my C: drive is no longer in the red.

First, download Ollama from: https://ollama.com/download

Creating a new folder: E:\ollama

Then run PowerShell as Administrator and navigate to the ollama folder：

cd E:\ollama

Alternatively, you can right-click the folder while holding Shift and select “Open in Terminal” or “Open PowerShell window here” if you’re using an older Windows version.

Set system environment variable OLLAMA_MODELS.

PS E:\ollama> setx OLLAMA_MODELS E:\ollama\models /M
成功: 指定的值已得到保存。

Install model at custom path:

PS E:\ollama> .\OllamaSetup.exe /dir=E:\ollama

.

Download model from https://ollama.com/ 

Restart ollama to make the OLLAMA_MODELS variable effective. Run the command in a
new terminal:

ollama run deepseek-r1:14b

He will download and run the model.