wildminder/ComfyUI-DyPE

Let us democratise high-resolution generation! (CVPR 2024)

[ECCV 2026] Official implementation of "HRDiT: Training-Free High-Resolution Image Generation with Off-the-Shelf Diffusion Transformer Models"

TBG Enhanced Tiled Upscaler and Refiner upscales up to 200MP with precise control. It features dual-model processing (structure + detail), per-pixel denoising to control where creativity is applied, and per-tile and per-object prompts, denoise, and seeds. Advanced fast previews. New generative tile fusion seamless joining, even on high-denoise.

Official implementation of "SEGA: Spectral-Energy Guided Attention for Resolution Extrapolation in Diffusion Transformers" [NeurIPS 2026]

[ICLR 2025] Official Implementation of Meissonic: Revitalizing Masked Generative Transformers for Efficient High-Resolution Text-to-Image Synthesis

SUPIR upscaling wrapper for ComfyUI

Arbitrary scale latent upscaler for SDXL, Flux, Flux2, and Wan 2.1/Anima/Qwen Image. Supports fractional scaling up to 2.0x. Includes ComfyUI node and pretrained weights.

Z-Image, Flux2 Klein, Krea2 & SeedVR2 with a Gradio UI. Uses a built-in ComfyUI backend for speed and efficiency! [8GB+VRAM, 16GB+ RAM]

MagicNodes, it's a plug-and-play multi-pass "render-machine" for SD/SDXL models. Simple one-node start, expert-grade results. Core is ZeResFDG (Frequency-Decoupled + Rescale + Zero-Projection) and the always-on QSilk Micrograin Stabilizer, complemented by practical stabilizers (NAG, local masks, EPS, Muse Blend, Polish).