⦁ Multiply

Multiply [DVB] is exactly what the name says: a multiplication node. You give it a factor and a couple of numbers, it multiplies them and hands you the result twice - once as a FLOAT and once as a rounded INT. It's a two-node-in-one convenience that exists because in video-batch math you're constantly rescaling things, and you often need both representations of the answer.

The classic use is scaling frames by duration: if your frame set runs at 24 fps and you know you want 3 seconds of it, multiple of 24 times... well, you get the idea. Or doubling a batch size, or scaling a step count, or turning a per-frame value into a total. The pack pairs it with Divide [DVB] and the general-purpose Calculation [DVB] (which does arithmetic plus the whole Python math module), and Multiply is the one you grab when all you need is "times this."

One quirk worth knowing before you set values: the node multiplies both inputs and adds the results - (float × multiple) + (int × multiple). So if you only mean to scale the float, leave int at 0, and vice versa. Feed both and you get the sum of both scaled, which is occasionally exactly what you want (a frames component plus an offset component) but usually a surprise if you didn't plan it.

The inputs and outputs that matter

• multiple (default 2) - the factor everything gets multiplied by.
• float (default 0) - the float value to scale.
• int (default 0) - the integer value to scale.

Outputs:

• FLOAT - the precise result.
• INT - the result rounded to the nearest whole number.

Both are always produced, so you can grab whichever type the downstream node wants without a conversion node.

Installing

It's part of Dream Video Batches (Alt Key Project / Dream Project):

cd ComfyUI/custom_nodes
git clone https://github.com/alt-key-project/comfyui-dream-video-batches.git
cd comfyui-dream-video-batches
pip install -r requirements.txt

Or find "Dream Video Batches" in ComfyUI Manager, install, restart. No models.

Where people get burned

The "both inputs are added" behavior is the one to watch - if you set float and int to meaningful values and expect two separate results, you'll get one summed result instead. Also note the INT output is rounded, not truncated, which matters when the number feeds a frame count or batch size where an off-by-one changes the clip length. And if you ever need anything beyond multiply - division, powers, trig, min/max - that's Calculation [DVB], not this node; Multiply is deliberately, almost stubbornly simple.