I am sure when comfyui desktop for mac gets updated we will be able to use other workflows. In the meantime, here is an adaptation for mac users of the workflow from realrebelai user with my appreciation for his workflow and information. This works on a macbook M3 max pro taking around 3 to 4 times what it takes in a nvidia 4090. The mac heats up fast running these models (as with qwen and others) and you can fry an egg on the back of the laptop after creating 3 to 5 scenes in high definition so use it at your own risk and be careful.It includes two model loaders. Of course, bypass the one you are not using and connect the one you are using to the lora loader if using version 1.1. One model loader is for bf16 and one for any quantized model (Q5 k_m is my favorite although you can change it for your own favorite)Now that loras for z image are appearing, a version 1.1 is included with a lora loaderMac MPS does not love fp8 so use bf16 or quantized modelsLINKS TO FILES: Text Encoder<a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/z_image_turbo/tree/main/split_files/text_encoders">https://huggingface.co/Comfy-Org/z_image_turbo/tree/main/split_files/text_encoders</a>VAE<a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/z_image_turbo/tree/main/split_files/vae">https://huggingface.co/Comfy-Org/z_image_turbo/tree/main/split_files/vae</a>Bf16 model<a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/z_image_turbo/tree/main/split_files/diffusion_models">https://huggingface.co/Comfy-Org/z_image_turbo/tree/main/split_files/diffusion_models</a>quantized models<a target="_blank" rel="ugc" href="https://huggingface.co/jayn7/Z-Image-Turbo-GGUF/tree/main">https://huggingface.co/jayn7/Z-Image-Turbo-GGUF/tree/main</a>enjoy

Files with SHA256: c2247064a14a0401a73426f76029e08e13c1136f9e8071d6296c125e93dff6e7