Integer Input (fal)

This is the whole-number knob of the fal connector pack. It outputs a single INT, and like its String, Boolean, and Float siblings, its actual purpose is to become a named parameter in your workflow's API when the graph runs on fal.ai. It carries min, max, and step so you can bound the field, same as Float Input, just constrained to integers. Locally it's a number node. Exposed through fal, it's the labeled, range-limited integer field your API takes.

It's part of the ComfyUI-fal-Connector, fal's tooling for running a ComfyUI workflow on their serverless GPUs. Build the graph, "Save as fal format," POST it to fal.run/fal-ai/comfy-server, and every Integer Input becomes an overridable numeric field - {"seed": 12345, "steps": 30} in, image out.

What you'd wire it to

Integers are the whole-number controls of a diffusion graph, and the list is basically the greatest-hits of things people want to change per request: seed, sampler steps, image width and height, batch_size. Seed is the poster child - expose it as an Integer Input and a caller can pin a result to reproduce it, or send a fresh number to re-roll. Steps lets them trade speed for quality. Width/height lets one workflow serve multiple aspect ratios. Each of those you want callable gets its own Integer Input with a clear name.

The inputs that matter

Five required inputs; two you'll actually think about:

• name - the API parameter key. Default int_input; rename it (seed, steps, width) so the field is legible and multiple integers don't collide.
• number - the value, and the default when a caller sends nothing. Defaults to 0.
• min / max - the allowed range, defaulting to the full 32-bit integer span (about −2.1 billion to 2.1 billion). That's the right neighborhood for a seed but far too loose for steps or dimensions - cap steps at something like 1–60, and keep width/height to values your model actually handles.
• step - the increment on the widget, default 1. Bump it to 8 or 64 for dimension fields so callers land on model-friendly multiples instead of an odd super like 517.

The single output is an INT. Wire it into any integer input - a KSampler's seed or steps, an Empty Latent Image's width/height/batch_size, a loop count. Ordinary int wire from there.

Installing it

The connector ships with fal's bundle, per the README:

git clone --recursive https://github.com/badayvedat/ComfyUI-fal.git
cd ComfyUI-fal
pip install -r requirements.txt

Add your fal API key to fal-config.ini (optional Hugging Face token too). The README warns about the path: the file lives in ComfyUI-fal/custom_nodes/ComfyUI-fal-Connector/, not the repo root. Start with main.py.

Only want the node in an existing ComfyUI? ComfyUI Manager → search ComfyUI-fal-Connector, or cd ComfyUI/custom_nodes && git clone https://github.com/badayvedat/ComfyUI-fal-Connector, then restart. No model downloads, no heavy dependencies.

Where people get burned

Two integer-specific ones. First, dimensions that aren't multiples of 8 (or 64, depending on the architecture) - leave step at 1 and a caller can send 513, which many models reject or handle badly; set step and min/max to keep them on legal values. Second, seed range: if you clamp max too low, you kneecap the seed space and re-rolls start repeating - a seed field wants the full 32-bit range. Otherwise it's the pack-wide caveat: in a plain ComfyUI with no fal-config.ini and no fal-ai/comfy-server target, the API-parameter behavior isn't there; this is just a number node until it runs inside fal's pipeline, and remote auth failures trace to the config file's location or a missing key, not the node.