Install ComfyUI manually in an identifiable, recoverable Python environment

Purpose and scope

Plan around the supported interpreter, device backend and a separate working directory, not a single pip command for every machine. Desktop and Portable users should follow their own installation paths instead.

Source-supported context

ComfyUI and PyTorch installation documentation provide platform-specific choices. pip check validates declared requirements, not GPU kernels, model loading or successful workflow execution. Source 1 Source 2 Source 3

Before you start

1. Identify the operating system, GPU, available disk space and intended backend.

2. Obtain source from the project's official release route and select a supported interpreter.

3. Preserve an existing working environment rather than experimenting with unrelated nodes inside it.

Steps and checks

The sequence below is an editorial procedure based on the cited context, not a diagnosis already confirmed for your environment.

Step 1. Create a virtual environment in a separate directory and verify sys.executable. Application, model and output locations can be planned independently.

Step 2. Use the official PyTorch selector for the actual hardware and platform; a CUDA command in a tutorial is not appropriate for every Mac, AMD or CPU installation.

Step 3. Install this ComfyUI revision's requirements in the same environment, then check declarations and basic Torch imports and record versions.

Step 4. Start the core and test the author's minimal baseline. Introduce one necessary extension at a time only after the baseline works.

Commands and request examples

Use the environment and shell identified above. Paths and variables are examples to adapt; these commands/snippets were not executed for your machine. Create the environment only after choosing a supported Python, then run the activation example for your OS, not both. Verify with python -c "import sys; print(sys.executable)"; activation can be avoided by calling the environment executable directly.

python -m venv .venv

# Windows PowerShell
.\.venv\Scripts\Activate.ps1

# Linux / macOS
source .venv/bin/activate

Completion check

Interpreter path, core version, backend and working directory are recorded, the service starts and a small workflow completes before extension installation.

Limitations and cautions

This guide has not installed software on your machine. A virtual environment does not isolate drivers, kernels or all external tools. When activation is restricted, call the environment's interpreter directly rather than globally weakening script policy.

Original sources

• ComfyUI manual installation — original record checked 2026-09-21.
• PyTorch local installation — original record checked 2026-09-21.
• pip check — original record checked 2026-09-21.

Localization prepared: 2026-09-22. The source-check dates above were inherited from the existing article; this translation does not claim they were all rechecked today. No GPU run, minimum-memory measurement or runtime guarantee is supplied.

Related troubleshooting and guides

• Which Python does ComfyUI Portable use? Keep startup and package installation aligned
• pip ResolutionImpossible: resolve conflicting requirements, not the download speed
• torch.cuda.is_available() is False: inspect the build, driver and runtime context