Ollama

LLaMA (Meta) provides open-weight large language models for flexible local or hosted inference, enabling transparent, controllable LLM workflows across diverse environments.