LogicUtil_Uniform Random Int

ComfyUI is full of places where a whole number matters - seeds, list indices, step counts - and LogicUtil_Uniform Random Int is the node that rolls one for you, reproducibly. It picks an integer between min_val and max_val using random.Random(seed).randint(), so the same seed always gives the same integer. It's part of the LogicUtil subpack of ComfyUI-JDCN, and it's the natural companion to the pack's random choice node: one picks from a text list, this one rolls a number.

How it works

Three inputs, one output, no surprises:

• min_val and max_val (INT, both inclusive) - the range you're rolling in.
• seed - the reproducibility dial. Same seed, same number, forever.
• Output: INT, ready to feed into anything expecting a whole number.

Both bounds are inclusive, so min_val=1, max_val=3 can roll 1, 2, or 3. And there's the same fallback its float sibling has: if min_val is greater than max_val, it returns min_val instead of erroring or swapping. It won't crash on you, but it also won't do what you meant - check your order.

Why you'd reach for it

Two classic jobs. The first is generating seeds: wire the INT output into your KSampler's seed input (converted or directly if the socket accepts it), and every run gets a fresh seed while staying reproducible - change the seed input to get a new roll, leave it fixed to replay a run. That's the "lock the seed, change one variable" discipline applied in reverse: the random node is the variable you're changing.

The second is picking an index. If you've got a list of options being navigated by index (say, via the pack's String List to Combo), a random int is the dynamic index that makes the pick vary per run. Between this and the choice node you can cover both styles: pick from a list, or roll a number and index into one.

One honest caveat: this rolls uniformly, so every integer in the range is equally likely. If you want weighted odds, this isn't the node. And while it's random, remember it's seeded random - if your workflow suddenly stops varying, check whether something upstream is pinning the seed to a constant.

Installing it

It ships with ComfyUI-JDCN, so one install covers the whole pack. ComfyUI Manager: Install Custom Nodes → search "JDCN", install, restart. Or:

cd ComfyUI/custom_nodes
git clone https://github.com/daxcay/ComfyUI-JDCN.git
cd ComfyUI-JDCN
pip install -r requirements.txt

Only dependency is piexif; no models to download. It's a small two-person hobbyist pack (Daxton Caylor and Jerry Davos), and the random nodes are the pieces of it that earn a place in real workflows. Simple, seeded, and honest about what it does - a rare combination in this ecosystem.