🐳多数值比较
Feed it a pile of numbers, get back the one you meant
- result
- secondary_result
- info
This is pure plumbing, and plumbing is most of ComfyUI. MultiNumberCompare takes up to ten integers, runs one of eight operations across them, and hands back the answer plus a readable explanation. From the 🐳 Pond Nodes pack (comfy_Pond_Nodes), it lives under 🐳Pond/Tools and it's the sort of node you install for one batch workflow and then keep reaching for.
The real-world use case: you have a batch of seeds, a set of CFG values, or a stack of step counts computed by other nodes, and you need the biggest, the smallest, or the average to drive something downstream. Instead of squinting at five widgets and typing the winner in by hand, you wire them all in and let the graph decide. Change an input upstream and the answer recomputes everywhere it's used.
The inputs that matter
Two required settings:
- num_inputs - how many of the
int_1throughint_10slots actually appear. Set this first; it's the dial that keeps the widget list tidy. - output_mode - the operation:
max,min,median,average,sum,sorted_asc,sorted_desc, orrange.
Then the optional int_1 … int_10 values (integers, ±999,999) you're comparing. Only the slots up to num_inputs count.
Three outputs: result (the main answer - for sorted_asc/sorted_desc this is where the sorted list logic lands), secondary_result (the second value the mode produces), and info (a STRING explaining what it did, e.g. which mode ran over how many numbers). That info string is the quietly useful bit - it tells you how many inputs actually participated, which catches the classic mistake of forgetting to bump num_inputs when you add values.
Installing it
Identical to every node in this pack. ComfyUI Manager → search comfy_Pond_Nodes → install → restart, or:
cd ComfyUI/custom_nodes
git clone https://github.com/Pondowner857/comfy_Pond_Nodes
cd comfy_Pond_Nodes
pip install -r requirements.txt
It's pure Python standard library - the pack's heavyweight requirements.txt (ultralytics, onnxruntime, mediapipe, realesrgan…) is for the image nodes, so this one runs fine on a bare install. And for what it's worth, the README warns that the pack conflicts with comfyui_HiDream-Sampler (endless console scrolling) - filed away for the day your console goes haywire.
The take
Don't overthink it. If you want one number out of several and it's getting tedious to compute by hand every run, this is the node. The median and range modes are the least common but genuinely handy in batch-variation workflows, and having both the sorted output and the explanation string makes it a better debug tool than a bare "get max" node. Slightly boring, exactly the point - a value node should be boring.
Inputs (12)
| Name | Type | Default | Description |
|---|---|---|---|
| num_inputs | INT | 22–10 | — |
| output_mode | COMBO | max | 8 options: max, min, median, average, sum, sorted_asc, +2 |
| int_1opt | INT | 0-999999–999999 | — |
| int_2opt | INT | 0-999999–999999 | — |
| int_3opt | INT | 0-999999–999999 | — |
| int_4opt | INT | 0-999999–999999 | — |
| int_5opt | INT | 0-999999–999999 | — |
| int_6opt | INT | 0-999999–999999 | — |
| int_7opt | INT | 0-999999–999999 | — |
| int_8opt | INT | 0-999999–999999 | — |
| int_9opt | INT | 0-999999–999999 | — |
| int_10opt | INT | 0-999999–999999 | — |
Outputs (3)
| Name | Type | Description |
|---|---|---|
| result | INT | — |
| secondary_result | INT | — |
| info | STRING | — |