Z-Image Power Nodes v1.0 has been released! A new version of the node set that pushes Z-Image Turbo to its limits.
New node set for ComfyUI delivers photorealistic images in 7-9 steps without LoRAs or post-processing.
Independent developer FotografoVirtual has launched Z-Image Power Nodes v1.0, a major update to the specialized node collection for the Z-Image and Z-Image Turbo models within the popular ComfyUI workflow system. The release is designed to push the Z-Image Turbo model to its limits, focusing on achieving production-ready image quality directly in the generation phase. The headline feature is a new, specialized sampler that enforces strict prompt adherence while delivering output detailed enough to bypass traditional post-processing steps. This is complemented by a massive library of 120 visual styles—50 of which are new—that can be applied to any prompt via a new, searchable Style Gallery dialog with previews.
The technical core of v1.0 is a complete code overhaul of the Z-Sampler, introducing an improved denoising process that produces more vibrant colors and a wider brightness range. Crucially, this process is adjustable, with 0% reverting to the old behavior. New nodes include an advanced Z-Sampler Turbo variant for fine-tuning internal noise calibration, a 'My Top-10 Styles' node for quick access to favorites, and a 'VAE Encode' node specifically for soft inpainting workflows. The developer demonstrates that the nodes can generate photorealistic images in just 7 to 9 steps without using LoRAs or external editors, positioning it as a significant efficiency boost for AI image generation pipelines. The nodes are available via ComfyUI-Manager or manual GitHub installation.
- Adds 50 new visual styles for a total of 120, accessible via a new searchable Style Gallery with image previews.
- Features a completely rebuilt Z-Sampler with adjustable denoising for richer colors and brightness, eliminating need for post-processing.
- Enables generation of high-fidelity, prompt-adherent images in 7-9 steps without relying on LoRAs or external editing software.
Why It Matters
Dramatically streamlines the AI image workflow by delivering final-quality output directly from the model, saving time and technical overhead for creators.