Budget AITarget: 4GB - 6GB
Minimal VRAM Toolkit for 4GB-6GB GPUs (GTX 1650 / RTX 2060)
Tailored for resource-constrained laptops and budget workstations. These lightweight nodes avoid GPU memory bloat and enforce aggressive model weight offloading.
📦 Recommended Nodes in this Kit
1
View Specs →ComfyUI-GGUFby city96
Quantizes big weights to fit inside tight 4GB-6GB buffers
2
View Specs →ComfyUI-SaveMemby hnmr293
Forces garbage collection and VRAM release between generations
3
View Specs →sd-latent-upscalerby city96
Zero-memory mathematical latent upscaling
4
View Specs →ComfyUI-TiledKSamplerby blenderneko
Renders huge high-resolution images in 512x512 memory chunks
⚡ Quick Terminal Install
Run these git commands inside your `ComfyUI/` root folder to clone the essential components of this pack:
git clone https://github.com/city96/ComfyUI-GGUF custom_nodes/ComfyUI-GGUFgit clone https://github.com/blenderneko/ComfyUI_TiledKSampler custom_nodes/ComfyUI_TiledKSamplerWant these pre-installed with zero setup?
Deploy a ComfyUI workspace with popular custom node packages already compiled on RunPod or Vast.ai.