How to run DeepSeek AI models on a Laptop

In this video we show how to run some of the "smaller" models offered by DeepSeek AI on a consumer-grade laptop using a combination of Ollama, Open WebUI, & Docker. DeepSeek AI is an AGI Research Lab based in Hangzhou, China that was founded in 2023. While DeepSeek AI has consistently been releasing open source models since 2023, their latest batch of "R1" models (released in January 2025) took the world by storm after the accompanying research paper claimed they were beating leading AI models developed by top AI companies on several industry standard benchmarks. Enjoy!

00:00 Intro
00:41 Background info about R1-8B
01:28 Explaining distillation
02:00 The demo begins
02:10 Download and install Ollama
03:53 Downloading and install Docker
05:04 Running R1-8B with Ollama
05:25 Browing Ollama.com while R1-8B is downloading
08:18 Running Open WebUI with Docker
09:30 Shut off internet for privacy
10:20 Interacting with R1-8B
11:35 Outro

RESEARCH PAPERS

DeepSeek-V3 Technical Report - https://arxiv.org/pdf/2412.19437
DeepSeek-R1 - https://arxiv.org/pdf/2501.12948
DeepSeekMath - https://arxiv.org/pdf/2402.03300

Видео How to run DeepSeek AI models on a Laptop канала COMMAND