Ollama v0.40.0: MLX Runtime Default on Apple Silicon

Core summary

One sentence to understand this update

Ollama v0.40.0 now automatically runs supported model architectures on the MLX runtime by default on Apple Silicon devices, enhancing performance for Mac users.

Impact & opportunity

What this could mean

Apple Silicon users can expect improved out-of-the-box performance and efficiency for local LLM inference with Ollama, simplifying deployment and increasing accessibility.