D-Ogi/ComfyUI-Attention-Optimizer
Automatically benchmark and optimize attention in diffusion models. 1.5-2x speedup on RTX 4090.
This tool helps artists and designers using ComfyUI to generate images and videos much faster. It automatically tests different underlying technologies for handling image details and selects the quickest one for your specific computer hardware and chosen AI model. You provide your AI model in ComfyUI, and it outputs an optimized version that generates content up to 2-4 times faster, especially for high-resolution images or longer videos.
Use this if you are a digital artist or content creator using ComfyUI for generative AI and want to significantly speed up your image and video creation times without needing to understand complex technical details.
Not ideal if you are not using ComfyUI, or if your primary goal is to save memory rather than pure generation speed.
Stars
27
Forks
5
Language
Python
License
MIT
Category
Last pushed
Feb 09, 2026
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/D-Ogi/ComfyUI-Attention-Optimizer"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
Comfy-Org/comfy-cli
Command Line Interface for Managing ComfyUI
YanWenKun/ComfyUI-Docker
🐳Dockerfile for 🎨ComfyUI. | 容器镜像与启动脚本
Comfy-Org/ComfyUI
The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface.
runpod-workers/worker-comfyui
ComfyUI as a serverless API on RunPod
Acly/comfyui-tooling-nodes
Nodes for using ComfyUI as a backend for external tools. Send and receive images directly...