Imagine needing a powerful GPU just to cybersecurity experiment with AI. Ridiculous, right? Hackers don’t rely on expensive hardware—we make the most of what we have. In the world of cybersecurity, knowledge is power, and AI is the next big weapon. But what if we could run it right on our low end Kali Linux computer, even without a GPU?
Most AI models are resource-hungry, and codes are locked behind corporate-controlled cloud services that spy on everything we do. Take ChatGPT, for an example—completely closed-source and tied to a centralized system that logs every interaction. That’s not how real hackers work. We need tools that are lightweight, private, and fully under our control. Come on That’s why we prefer Linux over Windows.
This is where DeepSeek comes in—a fully open-source AI model you can run locally, without sending your data to some shady cloud server. Yes, it’s a Chinese-developed model, but unlike ChatGPT, it’s transparent and open for anyone to inspect. No hidden tracking, no corporate lock-in—just raw AI power, running on our own computer.
The best part? We don’t need a high-end GPU. Even an older PC (Corei5 4th gen we are using) can handle DeepSeek with the right setup. Get ready to take control and bring AI to our Kali Linux machine? That’s enough intro let’s get started. 🚀
Installing DeepSeek on Kali Linux
We don’t have large computer systems with GPU we just got a simple Intel i5 4th gen CPU with 8 GB DDR3 RAM (It’s a 12 years old infrastructure). We run Kali Linux on it seamlessly, now on the Kali Linux system we are going to install DeepSeek on it.
To do that we need Ollama services on our system. To install and activate Ollama services on our system we need to run following command on our terminal:
curl -fsSL https://ollama.com/install.sh | sh
The above command will download & run the installation file for Linux environment as we can see in the following screenshot:
On the above screenshot we can see that the downloading process is started. It will take some time depending on our system resources and internet speed.
After downloading the bash script it will start installing that and then it will automatically starts ollama services. Everything is automatic. By any chance things go different we can check if everything is correct.
We can verify the installation by running following command:
ollama --version
Also we can check if Ollama is running or not by applying following command:
systemctl is-active ollama.service
If everything is set we can see Ollama version and active status as the following screenshot:
Now we start the main thing. Let’s install DeepSeek. As we are in low end system we are going to install a basic version of DeepSeek, to do so we need to run the following command:
ollama run deepseek-r1:1.5b
Here we choose 1.5b model, it’s a basic model for smaller scale AI works. This will run on our system perfectly. For larger systems owners here is the list.
Model | Download Size | RAM Needed | CPU | Best For |
---|---|---|---|---|
1.5B | ~1.1GB | 4GB | Old CPUs | Low-end PCs, Basic AI |
7B | ~4.7GB | 8GB | Mid-range CPUs | Balanced AI tasks |
8B | ~4.9GB | 10GB | Fast CPUs | Better AI performance |
14B | ~9GB | 16GB | Fast CPUs | Advanced AI tasks |
32B | ~20GB | 32GB | High-end CPUs | AI developers |
70B | ~43GB+ | 64GB | Workstation CPUs | AI research |
671B | ~404GB+ | 1.5TB | Workstation CPUs | AI research |
We can choose the model as per our system. For this tutorial and as per our system we have choose model 1.5b (shown in above command). Now it’s time to hit and run the above command. The screenshot shows the output of the above command.
Now we have DeepSeek on our system’s terminal. Let’s ask it something. Our chatting screenshot is following.
VoilĂ . We installed DeepSeek on our Kali Linux system locally which runs on very low end CPU. That’s how can easily install DeepSeek on Kali Linux or any other Debian based Linux system easily.
Enjoy our articles? Make sure to follow us on Twitter and GitHub, we post article updates there. To join our KaliLinuxIn family, join our Telegram Group & Whatsapp Channel. We are striving to build a community for Linux and cybersecurity. For anything we always happy to help everyone on the comment section and on Telegram Group. As we know our comment section is always open to everyone. We read each and every comment and we always reply.
Source: Read More