ComfyUI Extension: Resource Monitor

Authored by PBandDev

Created

Updated

1 stars

Run ComfyUI workflows without the setup

No installs, no CUDA version roulette, no GPU sitting idle on your bill. Bring a workflow and run it in the browser.

A ComfyUI-native live resource monitor and snapshot node

Looking for a different extension?

Custom Nodes (1)

README

comfyui-resource-monitor

comfyui-resource-monitor adds a simple live resource monitor to ComfyUI and a Resource Snapshot node for workflows. Created with comfyui-custom-node-template.

<img src="assets/top.png" alt="Resource Monitor Preview">

It is designed to feel native to ComfyUI:

  • Top mode mounts into the workflow action bar
  • Bottom mode mounts into the canvas toolbar
  • Collapsed mode becomes a Resources button in the top bar
  • styling follows ComfyUI's existing colors and UI chrome

The monitor can show:

  • CPU usage
  • RAM usage
  • GPU usage
  • VRAM usage
  • GPU temperature

Install

ComfyUI Manager

  1. Open Custom Nodes Manager
  2. Search and install comfyui-resource-monitor
  3. Restart ComfyUI

Use

After install, open ComfyUI settings and search for Resource Monitor.

Available settings:

  • Display Mode: Top, Bottom, or Collapsed
  • Expanded By Default
  • Refresh Rate
  • Smooth Transitions
  • Text Density: Compact or Detailed
  • Show Clear Buttons: show or hide the Unload Models and Free Memory toolbar buttons
  • metric toggles for CPU, RAM, GPU, VRAM, and GPU temperature

Changes apply immediately in the UI.

Clear Controls

Two toolbar buttons appear adjacent to the resource monitor:

  • Unload Models — unloads currently loaded models from VRAM
  • Free Memory — frees all memory, models, and VRAM

Resource Snapshot Node

The node appears under PBandDev/Resource Monitor as Resource Snapshot.

It captures the machine's current resource state at the point where the node runs and outputs:

  • passthrough
  • cpu_percent
  • ram_percent
  • ram_used_bytes
  • ram_total_bytes
  • gpu_percent
  • vram_percent
  • vram_used_bytes
  • vram_total_bytes
  • gpu_temp_celsius
  • gpu_available
  • gpu_name

The optional passthrough input is forwarded to the passthrough output so the node can sit in the middle of a workflow chain.

If GPU telemetry is unavailable, GPU-related numeric outputs fall back to 0 and gpu_available is false.

<img src="assets/node.png" alt="Resource Monitor Node Preview">

Troubleshooting

  • If the monitor does not appear, restart ComfyUI and reload the browser tab.
  • If the node appears but GPU values are 0, your system may not expose NVML telemetry.
  • If you changed frontend files locally during development, reload the browser tab.
  • If you changed Python files locally during development, restart ComfyUI.

Development

Install dependencies:

pnpm install
uv sync --locked --group dev

Checks:

uv run pytest tests/python -q
pnpm test
pnpm typecheck
pnpm build

Run ComfyUI workflows without the setup

No installs, no CUDA version roulette, no GPU sitting idle on your bill. Bring a workflow and run it in the browser.

Learn more