DEV Community

Avnish
Avnish

Posted on

How to Install DeepSeek-R1 32B on Windows: System Requirements, Docker, Ollama, and WebUI Setup

DeepSeek-R1 32B System Requirements Docker, Ollama, and WebUI Setup

Component Minimum Requirement Recommended Requirement
GPU NVIDIA RTX 3090 (24GB VRAM) NVIDIA RTX 4090 / A100 (40GB+ VRAM)
CPU 8-core processor (Intel i7 / AMD Ryzen 7) 16-core processor (Intel i9 / AMD Ryzen 9)
RAM 32GB 64GB+
Storage 100GB SSD 1TB NVMe SSD
OS Windows 10/11 Windows 11
Docker Support WSL2 enabled WSL2 enabled

Installation Methods

I’ll provide three different installation methods:

  1. Using Docker (For easy containerized deployment)
  2. Using Ollama (For a simplified local installation)
  3. Using WebUI (For an interactive browser-based experience)

1️⃣ Installing DeepSeek-R1 32B Using Docker

Step 1: Install Docker

  1. Download Docker Desktop for Windows.
  2. Run the installer and enable WSL2 Backend during setup.
  3. Restart your system and verify installation with: 
   docker --version
Enter fullscreen mode Exit fullscreen mode

Step 2: Pull DeepSeek-R1 32B Docker Image

Run the following command to download the DeepSeek-R1 32B image:

docker pull deepseek/deepseek-r1:32b
Enter fullscreen mode Exit fullscreen mode

Step 3: Run the Docker Container

Start the DeepSeek model in a container:

docker run -d --gpus all -p 8080:8080 --name deepseek-r1-32b deepseek/deepseek-r1:32b
Enter fullscreen mode Exit fullscreen mode
  • -d → Runs in detached mode
  • --gpus all → Allocates all available GPUs
  • -p 8080:8080 → Maps container port to the local machine

Step 4: Access the WebUI

  1. Open your browser and go to: 
   http://localhost:8080
Enter fullscreen mode Exit fullscreen mode
  1. Start interacting with DeepSeek-R1 32B!

2️⃣ Installing DeepSeek-R1 32B Using Ollama

Ollama is an alternative that simplifies running LLMs locally.

Step 1: Install Ollama

  1. Download Ollama for Windows
  2. Run the installer and follow the setup instructions.

Step 2: Install DeepSeek-R1 32B Model

Once Ollama is installed, run the following command:

ollama pull deepseek-r1:32b
Enter fullscreen mode Exit fullscreen mode

This will download and prepare the model.

Step 3: Run DeepSeek-R1 Locally

Start DeepSeek-R1 in interactive mode:

ollama run deepseek-r1:32b
Enter fullscreen mode Exit fullscreen mode

Now, you can chat with DeepSeek directly from the terminal.

3️⃣ Running DeepSeek-R1 with a WebUI

If you prefer an interactive WebUI, follow these additional steps.

Step 1: Install a WebUI (text-generation-webui)

Clone and install a text-generation-webui:

git clone https://github.com/oobabooga/text-generation-webui.git
cd text-generation-webui
pip install -r requirements.txt
Enter fullscreen mode Exit fullscreen mode

Step 2: Start the WebUI with DeepSeek

Launch the WebUI and specify the DeepSeek model:

python server.py --model deepseek-r1:32b
Enter fullscreen mode Exit fullscreen mode

Now, you can access DeepSeek-R1 via WebUI at:

http://localhost:5000
Enter fullscreen mode Exit fullscreen mode

Which Method Should You Choose?

Method Best For Ease of Setup
Docker Running in an isolated container ⭐⭐⭐
Ollama Quick setup and local execution ⭐⭐⭐⭐⭐
WebUI Browser-based interaction ⭐⭐⭐⭐

If you have a powerful GPU, Docker or Ollama are great choices.

If you prefer browser access, go for the WebUI setup.

Top comments (0)

Read next

thescottyjam profile image

Just 'cause something throws an error, doesn't mean you need to catch it.

theScottyJam -

codewander profile image

It feels like you can't reach simplicity with REST nor graphql, when it comes to joins. With graphql, the resolvers get complicated. With REST, it's easier to join the data using sql/orm, but then there is more boilerplate serializing the joining data.

Anon -

kevin_k profile image

AWS Aurora DSQL for Django Developers: A Step-by-Step Guide

Kevin Kiruri -

shyam0118 profile image

[Boost]

Shyamalendu Nayak -