Ollama uses llama.cpp in the background. It's more of a tool to use models: you download models as if they were docker images, you have a centralized way to org

Ollama uses llama.cpp in the background. It's more of a tool to use models: you download models as if they were docker images, you have a centralized way to organize your models, you can create new models with `Modelfile`... It's the docker of AI.