Really cool Z-Image text encoder node which formats your prompts as the model was trained. Developer has it imitate an LLM conversation and in doing so the model responds much better (+ node has many LLM creative stylizing templates)
https://github.com/fblissjr/ComfyUI-QwenImageWanBridge

https://redd.it/1parzxf
@rStableDiffusion

Eyes by Z-Image Turbo
https://redd.it/1pawk66
@rStableDiffusion

Advanced Prompt Techniques for Stronger Artistic Effects.

https://redd.it/1pawt6o
@rStableDiffusion

600mm Prime Lens Lora for z-Image

https://redd.it/1pax02z
@rStableDiffusion

Camera angles comparison (Z-Image Turbo vs FLUX.1 Krea)

https://redd.it/1pb02yu
@rStableDiffusion