Ino Math Float
Arithmetic between two numbers without hunting for a calculator node
- float
- int
Sooner or later every workflow needs a little math - scale a strength by 1.5, halve a denoise value, square something because reasons. Stock ComfyUI has no dedicated arithmetic node, which means people rig together converters or do the arithmetic in their heads and hardcode it. Ino Math Float is the honest solution: two float inputs, pick an operation, get the result - plus the integer-truncated version for free.
It's part of ComfyUI Ino Nodes (nobandegani/ComfyUI-InoNodes), the 125+-node V3-schema pack, and sits in its tiny InoFloatHelper category alongside float-to-int and float-comparison nodes.
How it works
Three inputs:
float_a- first number.float_b- second number.operation- a dropdown with six choices:add,subtract,multiply,divide,modulo,power.
The node computes the result and hands you it twice:
float- the exact float result.int- the same value truncated to an integer (Python'sint()rounding, so it floors toward zero).
That int output is the sleeper feature. Feed it power with a small exponent and you get a handy integer, or just grab it when your downstream node wants an INT and you'd otherwise need a separate float-to-int node. One compute, two typed outputs - that's the kind of frugality that keeps a graph small.
The gotchas
Two edge cases, both handled defensively rather than crashing:
- Divide by zero returns
0.0, and modulo by zero returns0.0too. If your divisor can legitimately be zero, the node won't explode - but it also won't tell you it silently gave you zero, so check your math assumptions before relying on the result. intoutput truncates. If you need rounding (up or to nearest), pair this with the pack'sIno Float To Intwhich offers round/floor/ceil instead of the truncation you get here.
When you'd use it
The bread-and-butter case is scaling: take a strength or CFG value, multiply it by a factor from a constant or another node, and feed the scaled number downstream. It's also handy for building dynamic sampler parameters - halve a step count, compute a denoise multiplier - where hardcoding numbers makes the workflow brittle.
Installing it
ComfyUI Ino Nodes installs like any custom node pack:
- ComfyUI Manager (easiest): search for "ComfyUI Ino Nodes", install, restart.
- Manual:
Then restart ComfyUI.cd ComfyUI/custom_nodes git clone https://github.com/nobandegani/ComfyUI-InoNodes.git cd ComfyUI-InoNodes pip install -r requirements.txt
The pack needs inopyutils and a current ComfyUI. It's a niche pack with minimal community footprint, so the README is your manual - but for "multiply two numbers and get a usable result," Ino Math Float is exactly the node you were looking for.
Inputs (3)
| Name | Type | Default | Description |
|---|---|---|---|
| float_a | FLOAT | 0.00 | — |
| float_b | FLOAT | 0.00 | — |
| operation | COMBO | 6 options: add, subtract, multiply, divide, modulo, power |
Outputs (2)
| Name | Type | Description |
|---|---|---|
| float | FLOAT | — |
| int | INT | — |