<

Less-than is the unglamorous half of the comparison family, and that's fine - you still need it every time the question is "is this under the cap?" Is the current step count below the threshold where we switch samplers? Is this resolution smaller than the tile limit? < with a node wrapper, from the Basic data handling pack, and it does exactly what it says.

What it does

Takes two numeric values and returns True if the first is less than the second, False otherwise. Plain Python value1 < value2. The boundary matters: 5 < 5 is False, so if you want "5 or less" you pair this with its <= sibling or use the pack's LessThanOrEqual node. Strict less-than is the right tool for "strictly under," and nothing else.

Like the rest of this pack's comparison nodes, both value1 and value2 accept FLOAT,INT - feed it a float from one side and an int from the other without casting. It's a small mercy you'll appreciate when your graph is full of mixed number types.

Inputs and outputs

• value1 - the first number (FLOAT or INT), the "under" side.
• value2 - the second number (FLOAT or INT), the upper bound.
• Output result - BOOLEAN, True when value1 < value2.

No modes, no options. Wire result into an if/else, a switch, or any boolean input.

Install

From the Basic data handling pack. Manager: search "Basic data handling" → install → restart. Manually:

cd ComfyUI/custom_nodes
git clone https://github.com/StableLlama/ComfyUI-basic_data_handling

then restart ComfyUI. The pack has zero runtime dependencies and downloads no models, so this is one of the safest custom-node installs around.

Worth knowing

If you've got both < and > nodes in a workflow, you can collapse the pair into the pack's NumberInRange node when you're testing an interval - one node, two bounds, no boolean algebra to keep straight. And as with every comparison here, cast text values to INT before comparing; Python compares strings alphabetically, and "2" < "10" is False in a way that will ruin your day.