β Int Input
A typed integer constant, because sometimes a primitive just isn't clear enough
- INT
This is a node that does literally nothing except hold an integer and hand it to whatever it's connected to. Int Input [Dream] takes a value, returns the same value. That's the entire implementation - in the source it's a noop. And yet it exists because the pack's author wanted a dependable, typed way to feed constants around a graph.
It's a utility node from the Dream Project Animation Nodes pack (alt-key-project), the 2023 Deforum-style animation toolkit. Unmaintained now, but a pass-through integer is the most stable kind of code there is.
What it is
One input: value (INT, default 0). One output: INT. The pack's own node list describes it as an "Integer input (until primitive routing issues are solved)" - that's the whole motivation. On the ComfyUI versions this pack targeted, the built-in Primitive node's routing was a known pain point, so the author shipped a family of explicit input nodes (Int, Float, String, Text) as a workaround.
When you'd use it (and when you wouldn't)
Honestly, on a modern ComfyUI you can usually reach for a primitive or just type the number into a widget. But this node still has two legitimate moments:
- Graph clarity. A named Int Input at the top of a workflow reads like a labeled constant: "fps" here, "frame_skip" there. When you share a workflow, that's friendlier than a bare primitive dangling mid-graph.
- Compatibility. If you're loading older workflows built around the Dream pack, the Int Input nodes are part of that graph's expectations - knowing what they do means the workflow isn't a mystery.
It also plays naturally with the pack's other int-hungry inputs: frame counters, offsets, multiplier knobs. Feed a single Int Input into several places and you have a one-point "change this number everywhere" control.
Installing
cd ComfyUI/custom_nodes
git clone https://github.com/alt-key-project/comfyui-dream-project.git
Or via ComfyUI Manager under "Dream Project Animation". No model downloads; the pack's deps are imageio, scipy, torchvision, pilgram, and evalidate - with the numpy<2.0 pin that's worth knowing about if you run into NumPy conflicts. Find it under Dream β utils.
Inputs (1)
| Name | Type | Default | Description |
|---|---|---|---|
| value | INT | 0 | β |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| INT | INT | β |