Ollama’s MLX engine has been updated to deliver its highest performance on Apple Silicon yet. Models output higher quality responses, respond faster, and use less memory.