Model
Trends
.ai
Model
Trends
.ai
les meilleurs modèles d'IA open source

Nœud Load Upscale Model dans ComfyUI

Load Upscale Model (UpscaleModelLoader) loads an ESRGAN-style upscaler such as 4x-UltraSharp or RealESRGAN from models/upscale_models for use with Upscale Image (using Model).
Pack de nœuds : ComfyUI core
Catégorie : Agrandissement
Cette page n'est pas encore traduite ; elle est affichée en anglais.

Ce que fait Load Upscale Model

Pixel upscalers are small neural networks that enlarge an image by a fixed factor (2x, 4x, sometimes 8x) and sharpen it as they go. This node loads one from the models/upscale_models folder and outputs an UPSCALE_MODEL object that the Upscale Image (using Model) node applies to an IMAGE.
Supported formats are the usual .pth and .safetensors files from ESRGAN, RealESRGAN, SwinIR, HAT, DAT and similar architectures (anything spandrel can read). The factor is read from the file, not chosen here.

Entrées

Nom
Type
Rôle
model_name
COMBO
The upscaler file from models/upscale_models.

Sorties

Nom
Type
Rôle
UPSCALE_MODEL
UPSCALE_MODEL
The loaded upscaler for Upscale Image (using Model) or Ultimate SD Upscale.
AD
Oubliez l'installation : des workflows ComfyUI préinstallés sur le cloud
BitVector exécute des workflows ComfyUI prêts à l'emploi sur ses propres GPU. Rien à installer, aucun nœud manquant, aucune case rouge. Ouvrez-le sur votre téléphone ou votre portable et générez en une minute.

Comment utiliser Load Upscale Model

  1. Download an upscaler (4x-UltraSharp, 4x_NMKD-Siax, RealESRGAN_x4plus, 4x-ClearRealityV1) into models/upscale_models.
  2. Add Load Upscale Model and pick it.
  3. Add Upscale Image (using Model), connect the UPSCALE_MODEL and the decoded IMAGE.
  4. Optionally add Upscale Image (scale by 0.5) afterwards to land on 2x.

Réglages et conseils

  • 4x-UltraSharp is a safe default for photos and renders; 4x_foolhardy_Remacri and 4x-AnimeSharp suit illustrations.
  • A 4x model turns 1024x1024 into 4096x4096 in one step; downscale afterwards if you only wanted 2x.
  • Pixel upscaling adds no new detail. For real detail, follow it with a low-denoise sampling pass (hires fix) or Ultimate SD Upscale.
  • Upscalers run on the GPU; big inputs can take a lot of VRAM, use ImageScale to shrink first if needed.

Dépannage de Load Upscale Model

Error: Unsupported model type / spandrel could not load

Pourquoi cela arrive
The file is not an image upscaler (for example a checkpoint, a face restoration model or a corrupt download).

Comment le corriger
Download the upscaler again from its original source and make sure it is an ESRGAN-type .pth or .safetensors file. Update ComfyUI so its spandrel copy knows newer architectures.

The dropdown is empty

Pourquoi cela arrive
There are no files in models/upscale_models.

Comment le corriger
Create the folder if it is missing, add an upscaler file, press R to refresh.

Out of memory during Upscale Image (using Model)

Pourquoi cela arrive
The upscaler processes the full image at once; 4x on a 2048 image needs a lot of VRAM.

Comment le corriger
ComfyUI already tiles automatically when it runs out of memory; if it still fails, downscale the input first or use a 2x model.

AD
Lancez ce workflow depuis votre téléphone
Chaque workflow de cette page est préinstallé sur BitVector, modèles et nœuds personnalisés déjà en place. Choisissez-en un, tapez un prompt, c'est fait. Zéro configuration, rien à télécharger.

Questions sur Load Upscale Model

Can I choose the upscale factor?

No, it comes from the model. Use a 2x model or follow with Upscale Image (scale by) to adjust.

Is this the same as latent upscaling?

No. This works on pixels after decoding. Upscale Latent works on the latent before a second sampler pass.

Nœuds liés

Upscale Image (using Model)
ComfyUI core
Upscale Image (using Model) runs an ESRGAN-style upscaler (4x-UltraSharp, RealESRGAN, 4x_NMKD-Siax, 4x_foolhardy_Remacri) over an IMAGE and returns it enlarged by the model's fixed factor with added detail.
Upscale Image and Upscale Image By
ComfyUI core
Upscale Image (ImageScale) and Upscale Image By resize an IMAGE with classic interpolation (nearest, bilinear, bicubic, lanczos, area) to an exact size or by a factor, with no AI model involved.
Upscale Latent and Upscale Latent By
ComfyUI core
Upscale Latent resizes a latent to a new size (or by a factor) so a second KSampler can add detail at higher resolution: the classic hires fix.
Ultimate SD Upscale
ComfyUI_UltimateSDUpscale
Ultimate SD Upscale enlarges a picture with an ESRGAN model, then re-samples it tile by tile through the diffusion model with your prompt at low denoise, adding real detail at 2x-4x without running out of VRAM.
AD
Marre de réparer des nœuds ? Laissez faire le cloud
BitVector garde des centaines de workflows ComfyUI installés, à jour et testés sur des GPU cloud rapides. Pas de Python, pas d'erreurs CUDA, pas de limite de VRAM. Depuis n'importe quel navigateur.

Autres nœuds ComfyUI

CLIP Text Encode (Prompt)
ComfyUI core
CLIP Text Encode turns a text prompt into CONDITIONING using the model text encoder. One node holds the positive prompt, a second one the negative prompt.
ComfyUI Manager
ComfyUI-Manager
ComfyUI Manager is the extension that installs, updates and fixes custom node packs and models from inside the interface, resolves missing nodes in imported workflows and snapshots your setup.
KSampler
ComfyUI core
KSampler runs the denoising loop: it takes the model, prompts and a latent and produces the finished latent image, controlled by seed, steps, cfg, sampler, scheduler and denoise.
Load Checkpoint
ComfyUI core
Load Checkpoint (CheckpointLoaderSimple) opens a .safetensors or .ckpt model file and hands out the three parts every workflow needs: the diffusion MODEL, the CLIP text encoder and the VAE.
Save Image
ComfyUI core
Save Image writes the IMAGE tensor to ComfyUI/output as a PNG, with the whole workflow embedded in the file metadata so the picture can be dragged back into ComfyUI to restore the graph.
VAE Decode
ComfyUI core
VAE Decode converts the sampled LATENT into a pixel IMAGE with the VAE; VAE Decode (Tiled) does the same in tiles for very large images.