🔶 Number Int

Nothing fancy here, and that's fine - sometimes you just need a plain integer sitting on its own node so you can see and adjust it without hunting through a widget buried inside another node. This is Chaosaiart's version of that: a single number_int field (default 1, range 0 to 99,999,999, step 1) that outputs an INT.

Where this actually earns its place in the pack is wiring. A lot of Chaosaiart's cache system runs on integer "reloader" slot numbers - chaosaiart_reloadAny_Save/Load, chaosaiart_reloadIMG_Save/Load, and chaosaiart_reloadLatent_Save/Load all take a reloader field as their cache slot ID. If you're building a workflow with several cache pairs and want to keep their slot numbers visibly consistent and easy to change from one place, routing them through a shared chaosaiart_Number node beats typing the same number into six different node widgets and hoping you don't fat-finger one of them. It's also just a convenient way to expose any integer value - a step count, a frame number you're testing against, whatever - as its own visible, labelable node in a complex graph, rather than a value buried three nodes deep.

There's genuinely nothing more to the mechanism than that: it holds a number, it outputs that number, every time the workflow runs. It doesn't know about frames, doesn't read the pack's restart/cache signals, and doesn't do anything context-dependent - for that you want chaosaiart_Number_Switch, which outputs a different value depending on whether it's the first run or a later one.

It's easy to underrate a node this plain, but it's the kind of thing that pays off specifically in a graph as sprawling as a full Chaosaiart animation setup tends to get - six cache pairs, a restarter, a multi-stage sampler chain, and a handful of frame prompts all on one canvas. At that scale, a labeled, single-purpose number node you can spot at a glance beats digging through widget after widget trying to remember which one you're supposed to be editing this time.

Install: search "Chaosaiart-Nodes" in ComfyUI Manager and install, or use Manager's "Install via Git URL" with https://github.com/chaosaiart/Chaosaiart-Nodes. By hand: cd ComfyUI/custom_nodes && git clone https://github.com/chaosaiart/Chaosaiart-Nodes, then restart ComfyUI. On Linux, also run pip install opencv-python and pip install tqdm inside your venv if you're using one - the pack's Windows installer handles this step automatically, the Linux path doesn't. No model downloads needed for this node or the pack in general; it just wires up whatever checkpoints, LoRAs, and VAEs you've already got loaded.

Not much to troubleshoot with a node this simple - if the output isn't what you expect, it's almost certainly because you're reading a different chaosaiart_Number node than the one you edited, which is an easy mistake once you've got several of these scattered around a busy graph. Give them distinct titles if you're using more than one, since the node itself doesn't label what the number is "for."