Ollama Has a Free API — Run LLMs Locally with One Command

Ollama runs large language models locally on your machine.

One command downloads and runs models including Llama 3, Mistral, Gemma, Phi, and 100+ others (e.g., 'ollama run llama3.1' auto-downloads).

It provides an OpenAI-compatible API at localhost:11434 with chat completion, generate, and tags endpoints.

GPU acceleration is supported for NVIDIA, AMD, and Apple Silicon.

Ollama is free and distributed under the MIT license.