floyo logo
Workflows
Pricing
floyo logo
Workflows
Pricing
Last updated
2026-06-12

Run hundreds of ComfyUI nodes and workflows in your browser.

ComfyUI and My Files on Floyo

ComfyUI-PiD is a set of custom nodes designed for use with NVIDIA's PixelDiT model, facilitating pixel diffusion decoding and upscaling workflows within the ComfyUI framework. This tool enables users to efficiently generate high-quality images from latent representations, leveraging advanced diffusion techniques.

  • Supports various NVIDIA models for enhanced image generation and upscaling.
  • Provides a streamlined workflow with nodes that automate tasks from text prompting to final image output.
  • Offers customizable settings for precision and quality control during the image generation process.

Context

ComfyUI-PiD is an extension for ComfyUI that integrates NVIDIA's PixelDiT model, specifically aimed at enhancing pixel-based image decoding and upscaling. This tool allows users to create detailed images from latent data, utilizing a series of specialized nodes that manage different stages of the image generation process.

Key Features & Benefits

The tool includes a variety of nodes that serve distinct functions, such as decoding latent images, generating captions, and performing upscaling. Each node is tailored to streamline the workflow, making it easier for users to produce high-quality outputs without needing to manually handle each step of the process.

Advanced Functionalities

ComfyUI-PiD features advanced capabilities, including support for multiple backbones and model precisions, which allow for flexibility in image generation. Users can select from different model settings to optimize for quality, speed, or specific output characteristics, tailoring the experience to their particular needs.

Practical Benefits

By utilizing ComfyUI-PiD, users can significantly enhance their workflow efficiency, gaining greater control over the image generation process. The tool's structured node system simplifies complex tasks, ensuring that users can focus on creative aspects while achieving high-quality results with minimal effort.

Credits/Acknowledgments

This project is maintained by contributors under the MIT license, with original authorship attributed to the developers of the ComfyUI and NVIDIA PixelDiT models.

Inner Nodes

PiDCaptionCreator
PiDDecode
PiDEmptyLatentImage
PiDFinalize
PiDKSamplerCapture
PiDPrepare
PiDSample
PiDTextPrompt
PiDUpscale

Discover most popular workflows

Hand-picked based on what hundreds of other artists looked at.