Boolean Input (fal)

This is the on/off switch of the fal connector pack. It outputs a single true/false, and - like its String, Float, and Integer siblings - its real purpose is to become a named parameter in your workflow's API when you run the graph on fal.ai. The name you give it is the JSON key a caller sets; the value is the default. In a plain local graph it's just a boolean constant. Exposed through fal, it's the toggle your API offers the outside world.

It's part of the ComfyUI-fal-Connector, fal's bridge for executing a ComfyUI workflow on their serverless GPUs. You build the graph, "Save as fal format," and POST it to fal.run/fal-ai/comfy-server; every Boolean Input turns into an overridable true/false field. That's how a fixed workflow becomes a configurable endpoint - {"upscale": true} in, result out.

What you'd wire it to

A boolean only matters if something downstream branches on it. On its own, ComfyUI doesn't do much with a bare true/false, so this node pairs with logic/switch nodes - the kind that route between two inputs, bypass a group, or pick one of two latents based on a condition. Think "hi-res fix on/off," "apply the LoRA or skip it," "return the upscaled image vs. the base." You declare the switch here so a caller can flip it per request, then feed the BOOLEAN into whatever gate acts on it. If your graph has no such gate, a Boolean Input is decorative - wire it to something that consumes it.

Inputs and outputs

Two required inputs:

• name - the parameter key in the API request. Default bool_input. Rename it to what it actually controls (enable_upscale, use_lora) so the field reads clearly to whoever calls the workflow, and so multiple booleans don't collide.
• value - the default, false out of the box. Set it to whatever the "if the caller says nothing" behavior should be.

One output: BOOLEAN. It goes into any node with a boolean input or a switch/conditional node. Nothing fancy - it's a clean true/false wire.

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 a fal API key to fal-config.ini (optional Hugging Face token too). The README warns that this file goes in ComfyUI-fal/custom_nodes/ComfyUI-fal-Connector/, not the repo root - an easy miss that breaks auth. Start the server with main.py.

If you just 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 - the whole connector is thin glue.

Where people get burned

The recurring confusion is expecting the node to do something by itself. It doesn't; it emits a value. If flipping it seems to have no effect, the branch you meant to control probably isn't reading the boolean - trace the wire to an actual switch. The other trap is the shared one across this pack: dropped into a plain ComfyUI with no fal-config.ini and no fal-ai/comfy-server target, the "API parameter" behavior simply isn't there - it's a local constant until it runs inside fal's pipeline. And if remote runs fail auth, check the config file's location and your key before suspecting the node.