This open-source model is rewriting the global rules of the AI game.

The newly released DeepSeek R1 inference model can actually compete head-to-head with OpenAI’s paid O1 model in professional fields like math problem-solving and code generation.

Even crazier? You can deploy it for free on your own computer.

Today, I’ll teach you how to quickly get this AI powerhouse up and running.

This varticle will also show you how to unlock intelligent features that are not yet officially supported.

Step 1: Download the Ollama client

Go to the Ollama official website and download the client version compatible with your system.

After installation, run the software and confirm that the little alpaca icon appears in your taskbar.

Step 2: Search for DeepSeek R1 on Ollama

On the Ollama official website, search for “DeepSeek R1.”

You’ll find multiple models of different sizes, optimized for various devices.

I’ve prepared a reference table to help you determine which model your computer can handle.

Step 3: Run the model

Running the model is straightforward.

Once you’ve selected your desired model, copy this magic command into your terminal:

When the progress bar fills up, congratulations! Your computer now has top-tier inference capabilities.

To test it out, type `/bye` to exit the model. Then enter `ollama list` to see other models available on your computer.

I switched to using the more advanced 14B version.

Step 4: Use ChatBOX for a better experience

While you can run the model directly in the terminal, it’s not very convenient.

I recommend using the ChatBOX browser extension.

Search for ChatBOX, launch the web version, and set the language to English.

Go to Settings and select “Ollama” as your model.

Follow this tutorial to ensure you can connect to your local service.

The setup process varies slightly depending on your operating system.

Once done, select the R1 model and save your settings.

Now you can enjoy a smooth experience using R1 right in your browser!

The real magic begins here: Create your own GPT-style AI

Click on “My Partner” and create a custom AI companion.

Define its personality and expertise, and voilà—a seasoned industry expert comes to life!

What you can do with local deployment？

Once installed locally, you can do so many cool things:

– Integrate it into Cursor for an offline, free AI coding tool.

– Develop automation tools or use it as a private knowledge base.

Best of all? It’s completely free and works offline, even without an internet connection!