EvaluateMultiple1

Sooner or later you hit the thing no dedicated node covers - a one-off calculation, a string reformat, a quick bit of conditional logic - and building a whole custom node for it is absurd overkill. That's what this node is for. The pack's own README describes the family in one blunt sentence: it exists to "support the execution of any fragment of Python code, generating multiple outputs from multiple inputs." EvaluateMultiple1 is the smallest member - one input, one output, whatever Python you write in between.

How it works

You type into the multiline expression box. in1 is whatever you wired into the optional input; your code reads it and is expected to produce out1. That's the entire contract: read in1, set out1. Unlike this pack's list-flavored sibling (EvaluateListMultiple), which ships with a hint pre-written in its expression box, EvaluateMultiple1 starts completely blank - you're writing from scratch, not editing a template.

Turn on print_to_console while you're first getting an expression right. It echoes whatever the node produced to your ComfyUI terminal, which is a far faster feedback loop than wiring a Show Text node just to check one value.

The input_count slider is worth knowing even here

input_count runs 1 to 20, and - notably - every one of the four EvaluateMultiple variants (1/3/6/9) ships that exact same range. The number in the node's name is just where it starts. Drag input_count up on this "1" node and it grows extra input/output sockets the same as the "3" or "9" node would have by default. So the honest answer to "which variant do I drop" is: whichever's closest, then dial input_count to fit - the separate sizes save you a few drags on the widget, they aren't differently capable.

The caution that actually matters

This runs real Python - no sandbox mentioned anywhere in the pack. Compare it to rgthree-comfy's Power Puter, a similarly-purposed expression node: its author was asked point-blank whether it's an arbitrary-code hole, and answered that input gets parsed to an AST and stepped through against an allowlist, so file access, network calls, and arbitrary imports are unavailable by design. RuiquNodes makes no such claim about EvaluateMultiple1 - the README says it runs "any fragment of Python code," full stop. Treat what you type here like a script you'd paste into a terminal: fine when it's your own logic, not something to copy blind from a workflow you downloaded without reading it first.

Installing it

ComfyUI Manager: search RuiquNodes, install, restart. Manual:

cd ComfyUI/custom_nodes
git clone https://github.com/ruiqutech/ComfyUI-RuiquNodes

then restart. No models or extra dependencies. Worth setting expectations: this is a small personal pack with essentially no footprint online - no tutorials, no help threads, nothing written about it beyond its own repo. If an expression misbehaves, you're debugging your own Python against the node's inputs and outputs, not chasing a known issue somewhere.

Common issues

A typo breaks the queue. The error will point at your expression, not the node's internals - read it like you would any Python traceback.

out1 comes back None when you expected a value. Check you're actually assigning to out1 inside the expression rather than just computing something and letting it fall on the floor - a bare expression with no assignment produces nothing to hand downstream.