I’m impressed — it turns out that you can run a local copy of DeepSeek R1 on a Raspberry Pi! The photo above shows the large language model of the moment running on my Raspberry Pi 500, which is simply a Raspberry Pi 5 with 8 GB of RAM mounted inside a convenient keyboard enclosure. DeepSeek R1 runs slowly on this machine, but it does run!
Fortunately, the installation process is pretty straightforward…
- Ensure the Raspberry Pi’s system is up to date by running the following commands in the terminal:
sudo apt update sudo apt upgrade -y
- Download and run the Ollama installer:
curl -fsSL https://ollama.com/install.sh | sh
- Download and start the DeepSeek model of your choice. I decided to see if the Raspberry Pi 500 could handle the 8b model, and it can!:
ollama run deepseek-r1:8b
The ollama run command first determines if you have the model specified after the command — in this case, the “8b” version of the DeepSeek R1 model, which is specified by deepseek-r1:8b.
If you don’t have it, it downloads the model and supporting files first. If you already have the model, it runs it.