How to use splats/gemma-4-26B-A4B-it-oQ4e with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir gemma-4-26B-A4B-it-oQ4e splats/gemma-4-26B-A4B-it-oQ4e