Ollama 0.30 is now available with improved performance and GGUF model compatibility through llama.cpp. This augments Ollama’s MLX engine on Apple silicon, bringing support to more models on a wider range of hardware.