⚠️ System Troubleshooting Guide

How to Fix "Torch not compiled with CUDA enabled" in ComfyUI

Replace CPU-only PyTorch with a GPU-accelerated CUDA build.

Error Signature:AssertionError: Torch not compiled with CUDA enabled

⚡ Quick Direct Answer

Reinstall PyTorch with explicit extra-index-url targeting your CUDA driver version (e.g., cu121 or cu124).

Why Does This Error Happen?

An automatic pip update or package installation installed the default CPU-only PyTorch wheel from PyPI, overwriting your CUDA-enabled GPU acceleration.

Recommended Terminal Repair Commands

Windows (ComfyUI Portable):

.\python_embeded\python.exe -m pip install torch torchvision torchaudio --force-reinstall --extra-index-url https://download.pytorch.org/whl/cu121

Linux / macOS (Virtualenv):

pip install torch torchvision torchaudio --force-reinstall --extra-index-url https://download.pytorch.org/whl/cu121

Still Crashing on Your Local GPU?

Skip compilation conflicts and memory limits by deploying on high-VRAM Cloud GPUs starting at pennies/hour.

Deploy on Vast.ai Deploy on RunPod

Frequently Asked Questions

How do I verify if CUDA is active in ComfyUI?

Open your ComfyUI terminal during startup. It should log "Total VRAM: ... MB, Device: NVIDIA GeForce...". If it says Device: CPU, CUDA is disabled.