Ino Get File Hash
Verify your downloads actually match
- success
- message
- hash
A corrupted model file is the worst kind of failure in ComfyUI: it doesn't error, it just silently produces garbage images and you spend an evening blaming the sampler. Hash checking is the fix, and most people never do it because the tooling feels like a detour into a terminal. Ino Get File Hash puts a SHA-256 check inside the graph - point it at a file, get the hash as a string, and compare it against the one the model page printed.
It comes from the Ino Nodes pack (nobandegani), in the file-helper category, and it follows the pack's standard file convention: parent_folder dropdown, folder, and filename inputs, plus the usual success/message reporting.
How it works
The node computes the SHA-256 hash of the file you name, reading it in chunks so it doesn't try to slurp a multi-gigabyte model into memory at once. The chunk_size input (in MB, default 8, range 1–1024) controls how big those reads are; for model files the default is fine, and you'd only raise it on very large files or slow disks where you want fewer, bigger reads.
Inputs to set:
- parent_folder - input, output, or temp, the ComfyUI root the file lives under.
- folder - subfolder path.
- filename - the file to hash.
- chunk_size - optional, MB per read chunk (default 8).
Outputs: hash is the 64-character SHA-256 hex string - that's the one you compare against a published checksum. success and message report whether the file was found and hashed.
Where it earns its place
Two genuinely useful spots. First, a download-integrity check: after a model download node writes a file, hash it and compare the result to the expected checksum. Wire the comparison into a switch or an abort condition and you've automated "is this model good to use" - genuinely valuable for big HuggingFace or Civitai downloads where a dropped byte is invisible until you generate. Second, de-duplication and change detection: hash a folder's files and identical content becomes obvious, which is exactly how the pack's own "Remove Duplicate Files" node thinks about files.
The honest caveat: hashing a 7 GB model takes a few seconds and some disk I/O - it's not free, so don't put it in a hot loop unless you actually need the verification.
Installing it
It ships with the whole Ino Nodes pack. ComfyUI Manager: search "ComfyUI Ino Nodes", install, restart. Or manually:
cd ComfyUI/custom_nodes
git clone https://github.com/nobandegani/ComfyUI-InoNodes.git
cd comfyui_ino_nodes
pip install -r requirements.txt
Restart after. Pack-wide caveats apply: it's built on ComfyUI's V3 schema (README wants v0.18.1+), and requirements.txt installs inopyutils, the helper library every node in the pack imports.
Common issues
If success comes back false, the message output tells you whether it's a missing file or a wrong path - check the parent_folder/folder/filename combo before anything else, since a path typo is the usual culprit. And make sure the hash you're comparing against is SHA-256: this node produces SHA-256 only, so a SHA-1 or MD5 checksum from a model page won't match no matter how correct the file is. Search "Ino Get File Hash" in the node menu (grouped under InoFileHelper).
Inputs (5)
| Name | Type | Default | Description |
|---|---|---|---|
| enabled | BOOLEAN | true | — |
| parent_folder | COMBO | 3 options: input, output, temp | |
| folder | STRING | — | |
| filename | STRING | — | |
| chunk_sizeopt | INT | 81–1024 | — |
Outputs (3)
| Name | Type | Description |
|---|---|---|
| success | BOOLEAN | — |
| message | STRING | — |
| hash | STRING | — |