How to Run DeepSeek Locally on Your Computer (PC) in 2025

Summary:

Want to Run DeepSeek locally on your computer (PC)? This guide covers everything—from setup on Windows, macOS, and Linux to running it on Android and iPhone. Learn how to use LM Studio, Ollama, and Open WebUI for a smooth experience. No internet, no data concerns—just full control over your AI assistant. Let’s set it up!

Ever wanted to run Deepseek AI (An AI chatbot like ChatGPT) on your own computer? No internet, no data tracking—just pure, private AI goodness? Well, DeepSeek R1 makes that possible, and trust me, it’s easier than you think.

In this guide, I’ll walk you through how to run DeepSeek locally on your computer (PC) using different methods. Whether you’re on Windows, macOS, Linux, or even a mobile device, I’ve got you covered. So grab a cup of coffee (or chai, if you’re in India), and let’s dive in!

Related: How I Applied a Screen Protector to My Samsung Galaxy S25 (Without Losing My Mind!)

Running DeepSeek R1 on Windows, macOS, and Linux

There are multiple ways to run DeepSeek locally, but I’ll focus on the most efficient ones: LM Studio, Ollama, and Open WebUI. Each has its own perks, so choose the one that suits you best.

1. Running DeepSeek R1 Using LM Studio

LM Studio is one of the easiest ways to get started. It’s a free, user-friendly app that lets you run AI models offline without needing a PhD in machine learning.

Here’s how to set it up:

• Grab LM Studio (version 0.3.8 or later) from the official site.
• Open it and head over to Model Search.
• Look for DeepSeek R1 Distill (Qwen 7B) in the list.
• Click Download and grab yourself a snack—it might take a few minutes.
• Make sure your system has at least 5GB of storage and 8GB of RAM.
• Once it’s downloaded, switch to the Chat tab.
• Select DeepSeek R1 as your model and hit Load Model.
• If you get an error, don’t panic—just set GPU offload to 0 in the settings.

And that’s it! You now have DeepSeek running offline, ready to answer your weirdest late-night questions.

2. Running DeepSeek R1 Using Ollama

If you prefer a more command-line approach (or just want to feel like a Pro), Ollama is a great option. It works on Windows, macOS, and Linux, and lets you run AI models with just a few commands.

• Download and install Ollama (it’s free!).
• Open Terminal (or Command Prompt if you’re on Windows).

Run this command to install the lightweight 1.5B model:
arduino
CopyEdit
ollama run deepseek-r1:1.5b 

If you have extra RAM and want better performance, go for the 7B model:
arduino
CopyEdit
ollama run deepseek-r1:7b 

• This version requires at least 4.7GB RAM, so make sure your system can handle it.
• Once installed, you can chat with DeepSeek directly from your Terminal.

To exit, just press Ctrl + D or close the window (no need to type any dramatic goodbye messages).

3. Running DeepSeek R1 Using Open WebUI

If you love the ChatGPT-style interface, Open WebUI is the way to go. It gives you a user-friendly chat window along with extra features like voice chat, file analysis, and a built-in code interpreter.

Here’s how to set it up:

• Install Python and Pip (if you don’t already have them).
• Open your Terminal and install Open WebUI using:

arduino
CopyEdit
pip install open-webui 

Start DeepSeek R1 with Ollama by running:

arduino
CopyEdit
ollama run deepseek-r1:1.5b 

Then, launch Open WebUI with:
arduino
CopyEdit
open-webui serve 

• Open your browser and go to http://localhost:8080.
• Boom! You now have a ChatGPT-style AI running offline.

To shut everything down, just right-click the Ollama icon and hit Quit, then close your Terminal.

Running DeepSeek Locally on Android & iPhone

If you’re more of a mobile user, don’t worry—I’ve got you covered. DeepSeek R1 works on both Android and iPhone using the PocketPal AI app.

Setting Up PocketPal AI

• Download PocketPal AI from the Google Play Store or Apple App Store.
• Open the app and tap Go to Models.
• Tap “+”, then choose “Add from Hugging Face”.
• Search for DeepSeek and select DeepSeek-R1-Distill-Qwen-1.5B.
• Download the model and load it.

Once done, you can chat with DeepSeek R1 directly on your phone—no internet required!

Minimum Requirements for Mobile Devices

Not all phones can handle AI models, so make sure your device meets these requirements:

• Android: Snapdragon 8 Elite, 8-series, or 7-series processor.
• iPhone: At least 6GB RAM for smooth performance.
• Storage: You’ll need at least 1.3GB free space.

Choosing the Right DeepSeek Model

DeepSeek R1 comes in different sizes, so here’s a quick breakdown:

Most AI models don’t use NPU (Neural Processing Unit) yet, so you’ll need a solid CPU and GPU for the best experience.

Why Run DeepSeek Locally?

So, why go through all this trouble when you could just use ChatGPT online? Here’s why:

• Privacy – No more data tracking or creepy AI spying.
• Works Offline – Use it anytime, anywhere—no Wi-Fi needed.
• Faster Responses – No waiting for servers to process requests.
• Customizable – Tweak it to fit your needs.
• No Subscription Fees – Say goodbye to expensive API costs.

Basically, if you want complete control over your AI experience, running DeepSeek locally is the way to go.

Troubleshooting Common Issues

If you run into problems, don’t worry—it happens to the best of us. Here are some quick fixes:

1. Model Not Loading in LM Studio?

• Set GPU offload to 0 in settings.
• Restart LM Studio and try again.

2. Ollama Not Working?

• Make sure it’s installed correctly.
• Update to the latest version.

3. Open WebUI Not Opening?

• Check if Python and Pip are installed.
• Run open-webui serve again.

4. Phone App Crashing?

• Make sure you have enough RAM and storage.
• Restart your phone and reload the model.

Related: Top 10 Useful Tips to Improve Samsung Galaxy Buds Battery Life

The Final Words

Running DeepSeek locally on your computer (PC) isn’t just a cool tech project—it’s a game changer. Whether you use LM Studio, Ollama, or Open WebUI, you get total privacy, speed, and customization.

So, go ahead, set it up, and enjoy the freedom of running your own AI. Who needs the cloud when you’ve got DeepSeek right on your machine?

Be a true Tech Enthusiast and get all Unplux publications update’s early access by joining our Telegram Channel and WhatsApp Channel.