Ollama

Ollama is a framework for running large language models locally, supporting Llama 3.3, DeepSeek-R1, Phi-4, Mistral, Gemma 2, and other models.