Run Deepseek R1 on your laptop in 5 minutes or less

Operating and hosting an LLM model like Deepseek locally allows you to explore and generate ideas using your local machine as a powerhouse, enabling tasks such as reasoning and building agents.

The simplest way to begin is by using Ollama, which offers direct access to quantized, distilled versions of models such as Deepseek, Qwen, Mistral, and others. Additionally, OpenWebUI provides a user-friendly interface for working with these LLM models locally.

Getting started

1. Download Ollama for your machine

   Install ollama

   Install Ollama

2. Install Deepseek R1 using Ollama with command

ollama run deepseek-r1

1. You can try ask question right away in your terminal

   Deepseek R1 in CLI with ollama

2. Now here is the fun part, download OpenWebUI, preferred is using Docker, once the open web ui container starts running visit http://localhost:3000/

   Using openwebui

   OpenWebUI

3. Sign up with your credentials, select the model which you installed in earlier steps and start asking !

   Deepseek R1 using OpenWebUI

Please note that the latency of the model responses depends on machine hardware. Additionally GPUs can also be used to accelerate response generations.

References :

• Ollama 

• GitHub - open-webui/open-webui: User-friendly AI Interface (Supports Ollama, OpenAI API, ...) 

• deepseek-v3 

• deepseek-r1