OTX Integer Multiple Inputs 4

Ever rebuilt a workflow and looked at the wall of little gray Primitive nodes - one for the seed, one for steps, one for the batch count, one for a Lora weight - and thought "there has to be a tidier way"? That's the entire pitch of OTX Integer Multiple Inputs 4. It's four typed integer slots in a single node, each settable by typing a number or wired from elsewhere, and each passed out untouched on its own output. That's it. No API, no GPU, no models.

It comes from Otonx's Custom Nodes, a small personal pack by Budi Hartono. Before you get excited: this pack has essentially no community footprint - you won't find it debated on reddit - and the README itself warns that nodes "are subject to intermittent changes. Please review and test nodes thoroughly prior to integration." For a node this dumb, that's a manageable risk.

How it works

The source is about as simple as a ComfyUI node gets. pass_parameters receives the four integers and returns them unmodified:

def pass_parameters(self, int_1, int_2, int_3, int_4):
    return int_1, int_2, int_3, int_4

There's no magic, and there's nothing to break. The value of the node is purely organizational: instead of scattering four single-value nodes across your graph, you keep them in one place where you can see and edit them together. Because the outputs are strictly typed INT rather than wildcard "any", the wiring stays honest - you can't accidentally shove a string into something that wants a number.

The inputs and outputs that matter

Every slot is the same shape: int_1 through int_4, each an INT with default 0, minimum 0, maximum 10000. Each has a matching INT output with the same name. Practical wiring:

• seed, steps, cfg multiplier, batch size - the classic four that every text-to-image run wants in one place
• image dimensions when your workflow takes fixed width/height
• A Lora stack that needs several integer-ish strengths in the same run

One honest limitation to know before you lean on it: the 0–10000 ceiling. If you're planning to feed big seeds (samplers happily take values up to 2^64), this box will quietly cap your seed choices. Use it for settings that live in that range - steps, batch, width, height - and let a proper seed node handle the big numbers.

Installing it

The whole pack is pure Python with zero dependencies - no requirements.txt, no model downloads, nothing heavy. Install via ComfyUI Manager (search "Otonx"), or clone it manually:

cd ComfyUI/custom_nodes
git clone https://github.com/budihartono/comfyui_otonx_nodes

Restart ComfyUI and you'll find it under the OtonxPack category in the node menu.

Common issues

• Can't find the node after install - restart ComfyUI fully. It registers as "OTX Integer Multiple Inputs 4" under OtonxPack, so search "OTX" if the category isn't showing.
• Values snap to 0 or refuse to go above 10000 - that's the widget's range doing its job, not a bug. If you genuinely need bigger integers, this isn't the node for it.

A quick honest take: if you already have four primitives wired and working, this won't speed anything up - it's a tidiness node, not a capability node. But for a workflow you rebuild often, four numbers in one box beats four boxes every time.