Deepseek-r1 is a revolutionary AI model out of China. In this short video, I show how to self-host Deepseek-r1 on an AWS T4 GPU using Control Plane and Ollama.

Control Plane makes it easy to run any containerized app across any cloud, or on-prem. In this video, I try (and succeed) to run Deepseek R1 so I can perform private LLM inference with this new powerful model.