Save 4 hours! We auto-setup your workflow! Free!

Drop your workflow.json — we handle every dependency, custom node, and model. Just open the link and run.

Auto-Setup Workflow Json (Free) Now!
ComfyUI > Nodes > Klein Tiled Upscaler

ComfyUI Extension: Klein Tiled Upscaler

Repo Name

ComfyUI_KleinTiledUpscaler

Author
gavr728 (Account age: 1145 days)
Nodes
View all nodes(1)
Latest Updated
2026-05-26
Github Stars
0.02K

How to Install Klein Tiled Upscaler

Install this extension via the ComfyUI Manager by searching for Klein Tiled Upscaler
  • 1. Click the Manager button in the main menu
  • 2. Select Custom Nodes Manager button
  • 3. Enter Klein Tiled Upscaler in the search bar
After installation, click the Restart button to restart ComfyUI. Then, manually refresh your browser to clear the cache and access the updated list of nodes.

Visit ComfyUI Online for ready-to-use ComfyUI environment

  • Free trial available
  • 16GB VRAM to 80GB VRAM GPU machines
  • 400+ preloaded models/nodes
  • Freedom to upload custom models/nodes
  • 200+ ready-to-run workflows
  • 100% private workspace with up to 200GB storage
  • Dedicated Support

Run ComfyUI Online

Klein Tiled Upscaler Description

Klein Tiled Upscaler is a high-fidelity tool designed to upscale images without running out of memory (zero-OOM) by using adaptive processing, specifically optimized for Flux.

ComfyUI_KleinTiledUpscaler Introduction

ComfyUI_KleinTiledUpscaler is an innovative extension designed to enhance the creative upscaling process for AI-generated images. This extension is particularly useful for AI artists who wish to add intricate details and textures to their images without simply enlarging the existing pixels. By utilizing an inpainting-based approach, ComfyUI_KleinTiledUpscaler allows the AI to generate new, high-frequency details such as organic textures and realistic grain, resulting in a more coherent and visually appealing image. This tool is especially beneficial for those working with the Flux2.Klein framework, offering a VRAM-friendly solution that can handle large-scale upscaling tasks even on budget GPUs.

How ComfyUI_KleinTiledUpscaler Works

At its core, ComfyUI_KleinTiledUpscaler operates by dividing an image into a grid of tiles and applying an inpainting process to each tile. This method differs from traditional upscaling techniques that often rely on mathematical algorithms to simply enlarge pixels. Instead, the extension uses AI to imagine and create new details within each tile, enhancing the overall texture and depth of the image. By processing each tile individually, the extension ensures that the final image is rich in detail and free from common issues like visible seams or mismatched boundaries.

ComfyUI_KleinTiledUpscaler Features

Symmetrical Canvas Partitioning

The extension automatically divides your image into equal-sized tiles, ensuring uniformity and eliminating issues with partial tiles at the edges. This feature is particularly useful for maintaining consistency across the entire image.

Built-In Reference Latents

ComfyUI_KleinTiledUpscaler handles reference latent extraction and structural guidance internally, streamlining the upscaling process and ensuring that each tile aligns perfectly with the original image.

Ground-Truth Laplacian Detail Analysis

This feature uses a GPU-accelerated Laplacian edge convolution to analyze the original image, distinguishing between flat and detailed areas. This prevents unwanted details from appearing in smooth regions like skies or walls.

Dynamic RoPE Alignment

The extension dynamically adjusts the Rotary Position Embedding coordinates to match each tile's location, ensuring seamless integration of details across the image.

VRAM-Friendly Architecture

By processing tiles sequentially, the extension minimizes VRAM usage, allowing for high-resolution upscaling even on less powerful hardware.

LoRa's Support

The extension supports LoRa models for Flux2.Klein, enabling additional customization and style consistency during the upscaling process.

ComfyUI_KleinTiledUpscaler Models

The extension is designed to work seamlessly with various models within the Flux2.Klein framework. While specific models are not detailed, users can leverage LoRa models for enhanced upscaling, consistency, and style adjustments. These models can be particularly useful for achieving specific artistic effects or maintaining a consistent look across multiple images.

Troubleshooting ComfyUI_KleinTiledUpscaler

If you encounter issues such as visible seams or color transitions, consider the following solutions:

  1. Enable Color Match: Set color_match to True to ensure consistent luminance and contrast across tiles.
  2. Increase Mask Blur: Adjust mask_blur to soften the boundaries between tiles.
  3. Use Detail-First Strategy: This setting prioritizes detailed areas, helping to anchor the image and blend smoother regions effectively.
  4. Adjust Adaptive Tiling: Experiment with turning adaptive tiling on or off to achieve the best results for your specific image.
  5. Utilize Consistency LoRas: These can help maintain uniformity across tiles, though they may limit the strength of the upscaling effect.

Learn More about ComfyUI_KleinTiledUpscaler

For further exploration and support, AI artists can access a variety of resources, including tutorials, community forums, and documentation specific to the Flux2.Klein framework. Engaging with these resources can provide valuable insights and tips for maximizing the potential of ComfyUI_KleinTiledUpscaler in your creative projects.

Klein Tiled Upscaler Related Nodes

RunComfy
Copyright 2025 RunComfy. All Rights Reserved.

RunComfy is the premier ComfyUI platform, offering ComfyUI online environment and services, along with ComfyUI workflows featuring stunning visuals. RunComfy also provides AI Models, enabling artists to harness the latest AI tools to create incredible art.

Klein Tiled Upscaler detailed guide | ComfyUI