NLP & Language · intermediate · concept 91 of 176
Open-Weights & Small Language Models
Models whose weights you can download, run, and tune yourself: the Llama, Qwen, DeepSeek, and Mistral lines. Small language models in the 1B to 14B range now do real work on laptops and phones. The open-versus-closed gap keeps narrowing, and pricing power follows it.
Key terms
Open weights vs open sourceSLMLocal inferenceModel licensesDistilled models
Learn these first
Where you meet it in the real world
On-device assistants, data-sovereign deployments, edge inference, fine-tuned vertical models
Videos
▶ Large Language Models explained briefly ↗
3Blue1Brown · YouTube
▶ Should You Use Open Source Large Language Models? ↗
IBM Technology · YouTube
Guides and articles
Courses, papers, and more