Ollama
Run open-weight LLMs and embedding models locally with one command, behind a simple HTTP API.
- #llm
- #local
- #embeddings
- #open-source
Ollama downloads and runs open-weight language and embedding models on your own machine, on macOS, Linux or Windows. It serves them through a local HTTP API, including an OpenAI-compatible endpoint.
Why it’s useful
- Free local embeddings: for prototyping a RAG pipeline, local embedding models cost nothing per token. Use the RAG Chunk & Token Calculator to compare that with hosted pricing at your data size.
- Private: your data never leaves the machine.
- One command:
ollama run <model>pulls the model and starts it.