How to Run DeepSeek R1 Locally on Windows, macOS, Android, & iPhone?

In the evolving world of artificial intelligence, the launch of DeepSeek R1 has created waves. Developed by a Chinese team, this model has rapidly surpassed ChatGPT in terms of user adoption, even claiming the top spot on the US App Store. As a result, DeepSeek R1 has caught the attention of AI enthusiasts, developers, and privacy-conscious individuals alike.

While DeepSeek R1 is free to use on its official website, many users have raised concerns about their privacy, especially considering the data storage in China. If you’re someone who values privacy and security, running DeepSeek R1 locally on your device is the ideal solution. In this guide, we’ll walk you through the different ways to set up DeepSeek R1 on your Windows PC, Mac, Android, and iPhone, ensuring that you maintain full control over your data while still harnessing the power of this advanced AI model.

Why Run DeepSeek R1 Locally?

Privacy is one of the primary reasons to consider running DeepSeek R1 locally. When you use cloud-based solutions, your data can be stored on external servers, which raises concerns about who has access to it. For instance, DeepSeek R1’s official platform stores data in China, which may not be acceptable for users in countries with strict privacy regulations.

By running DeepSeek R1 locally, you eliminate the risk of having your data shared with third parties. Local hosting also means that you won’t rely on an internet connection for AI-powered tasks, giving you a more seamless experience while ensuring complete privacy.

Additionally, running the AI model locally lets you customize your setup based on your hardware’s capabilities, whether you have a low-end laptop or a powerful desktop setup with high RAM and processing power.

Minimum System Requirements

Before diving into the installation process, let’s first look at the minimum system requirements you need to run DeepSeek R1 locally. Depending on your device and the size of the model you intend to run, these requirements may vary slightly:

1. PC, Mac, or Linux (Desktop Systems)

To run DeepSeek R1 on a desktop system, you need a computer with at least the following:

• Memory (RAM): A minimum of 8GB of RAM is required to run smaller models like the 1.5B. If you’re planning on running larger models like 7B or 14B, you may need up to 16GB or more.
• Storage: You’ll need enough storage space to accommodate the model files, which can vary depending on the model size (around 5GB to 8GB of space).
• Processor (CPU): While a fast processor isn’t a strict requirement, a multi-core processor will make things smoother, especially when working with larger models.
• Graphics Processing Unit (GPU): DeepSeek R1 uses CPU-based processing by default, but using a GPU (preferably an Nvidia GPU with CUDA support) can speed up computations, especially for larger models like the 32B or 70B versions.

2. Android Phones and iPhones (Mobile Devices)

On mobile devices, the requirements are a bit different but still manageable:

• Memory (RAM): A minimum of 6GB of RAM is recommended. This ensures smooth performance when running DeepSeek R1 on mobile devices.
• Processor (Chipset): Devices with Snapdragon 8 series (or higher) processors work best. Other high-performance chipsets like Apple’s A-series also support local AI models.
• Storage: Depending on the model you choose, you may need around 1GB to 2GB of available storage for downloading and installing the DeepSeek R1 model files.

How to Run DeepSeek R1 on Your PC or Mac

Now that you’re aware of the system requirements, let’s dive into the different ways you can run DeepSeek R1 locally on your Windows, macOS, or Linux-based PC. The methods we’ll cover include LM Studio, Ollama, and Open WebUI, each catering to different needs and preferences.

1. Using LM Studio (Windows, macOS, Linux)

LM Studio is one of the easiest ways to run AI models locally, including DeepSeek R1. It’s a free application that allows you to download and run a variety of AI models with ease, providing a user-friendly interface for beginners and advanced users alike.

Steps to Install LM Studio and Run DeepSeek R1:

• Step 1: Download and Install LM Studio
  • Visit the LM Studio website and download the latest version (0.3.8 or later).
  • Follow the installation instructions for your operating system (Windows, macOS, or Linux).
• Step 2: Search for DeepSeek R1 Model
  • Open LM Studio after installation.
  • In the left pane, go to the Model Search section and type in DeepSeek R1 Distill (Qwen 7B).
  • Select the model, which is based on the 7B version of DeepSeek R1.
• Step 3: Download the Model
  • Click the “Download” button. Make sure you have at least 5GB of storage available and 8GB of RAM to run this model smoothly.
• Step 4: Load the Model
  • Once the model is downloaded, navigate to the Chat window in LM Studio.
  • Select the model you just downloaded and click Load Model.
  • If you face any issues, try reducing “GPU offload” to 0 in the settings.
• Step 5: Start Chatting
  • After the model is loaded, you can now interact with DeepSeek R1 directly from your PC. Enjoy chatting or using the model for any AI-powered task!

2. Using Ollama (Windows, macOS, Linux)

Another great way to run DeepSeek R1 locally is by using Ollama. Ollama is a free tool that provides an easy-to-use command-line interface to run AI models, including DeepSeek R1.

Steps to Install Ollama and Run DeepSeek R1:

• Step 1: Install Ollama
  • Download Ollama from the official website.
  • Follow the installation instructions for your operating system.
• Step 2: Run the Model via Terminal
  • Open your Terminal (Command Prompt on Windows) and use the following command to run the 1.5B version of DeepSeek R1:arduinoCopyEditollama run deepseek-r1:1.5b
  • If your system supports it, you can run the larger 7B model with:arduinoCopyEditollama run deepseek-r1:7b
  • These commands will download the model and load it in your local environment.
• Step 3: Chat with DeepSeek R1
  • After running the command, you can chat with DeepSeek R1 directly from your Terminal window.
• Step 4: Exit the Chat
  • To exit the chat, simply press Ctrl + D.

3. Using Open WebUI for a ChatGPT-like Interface

If you prefer a graphical interface similar to ChatGPT, Open WebUI is the way to go. This option integrates with Ollama and gives you a web-based interface to interact with DeepSeek R1.

Steps to Install Open WebUI and Run DeepSeek R1:

• Step 1: Install Python and Pip
  • Before you start, ensure that Python and Pip are installed on your system.
  • Use the following command to install Open WebUI:arduinoCopyEditpip install open-webui
• Step 2: Set Up Open WebUI
  • After installation, open your Terminal and run the following command to start the Open WebUI server:arduinoCopyEditopen-webui serve
  • Open your web browser and go to http://localhost:8080.
• Step 3: Use DeepSeek R1 in a Web Interface
  • In the Open WebUI interface, you can select DeepSeek R1 from the dropdown menu.
  • Start chatting with DeepSeek R1 as you would in any other web-based chatbot interface.

How to Run DeepSeek R1 on Android and iPhone

Running DeepSeek R1 on mobile devices is possible, and with PocketPal AI, you can easily run local models without any cost. This is an excellent option for users who want AI functionality on the go without relying on internet connectivity.

1. Using PocketPal AI (Android and iPhone)

PocketPal AI is available for free on both Android and iOS. Here’s how you can run DeepSeek R1 on your mobile device:

Steps to Install PocketPal AI and Run DeepSeek R1:

• Step 1: Install PocketPal AI
  • Go to the Google Play Store (Android) or Apple App Store (iOS) and download PocketPal AI.
• Step 2: Add the DeepSeek R1 Model
  • Launch the PocketPal AI app and tap on Go to Models.
  • Tap the + button at the bottom-right corner to add a model from Hugging Face.
  • Search for DeepSeek R1 and find the DeepSeek-R1-Distill-Qwen-1.5B model by Bartowski.
  • Select the model that fits your device’s memory (typically the Q5_K_M model for phones with around 1.3GB of RAM).
• Step 3: Load the Model
  • Once downloaded, tap Load to initialize the model.
  • You can now chat with DeepSeek R1 locally on your Android or iPhone.

Final Thoughts

Running DeepSeek R1 locally offers a more secure and private way to interact with this advanced AI model. Whether you’re using it for creative writing, coding, or general inquiries, hosting it on your own device ensures that your data stays private. Depending on your device and hardware, you can choose the best method to install DeepSeek R1, whether it’s using LM Studio for a simple setup, Ollama for command-line enthusiasts, or Open WebUI for a more user-friendly experience.

For mobile users, PocketPal AI offers a fantastic solution to run DeepSeek R1 on both Android and iPhone, making it easier than ever to have this powerful AI in your pocket.

No matter which method you choose, you’ll be able to enjoy the full capabilities of DeepSeek R1 locally, with better control over your data and enhanced privacy.

Happy exploring with DeepSeek R1!