📊 Float Slider

Float Slider is exactly what it sounds like: a little node that gives you a draggable slider and outputs a single FLOAT. Its reason to exist is ergonomics - when you're dialing in a value that lives on a scale of 0 to 1, like a LoRA strength, a denoise, a ControlNet weight, or a blend factor, a slider you can nudge beats typing numbers into a field and re-running.

The mechanism is trivial and that's the point. The num input is the slider value, defaulting to a 0–1 range with a 0.01 step, and the float output is that value passed straight through to wherever you wire it. Where it gets a touch smarter is the optional settings input: a small JSON string, defaulting to {"min": 0.0, "max": 1.0, "step": 0.01}, that reconfigures the slider's range and granularity. Want a slider that runs 0 to 2 for an over-strength LoRA, or 0 to 10 with coarser steps for something else? Change the min/max/step in that JSON and the slider re-scales to match. So it's a slider you can retarget to whatever range the parameter downstream actually wants.

The input you care about is num - that's the thing you drag - and settings only if the default 0–1 range doesn't suit. The float output wires into any node that takes a FLOAT: a KSampler's denoise or CFG (converted to an input), a LoRA loader's strength, an IPAdapter weight, a resize scale. It's a control surface, nothing more.

Why bother over just typing the number? Two reasons. One, when you're tuning by feel - nudging a strength up and down to find the sweet spot - a slider is faster and more legible than editing a text field. Two, when you want one value to drive several nodes at once: put a Float Slider up front, fan its output into three different strength inputs, and now one drag moves all of them in lockstep. That's the genuinely useful pattern. For a workflow you re-run while eyeballing a result, having your key knobs as sliders at the top of the graph is a nicer way to work than hunting for the field buried in some node three columns over.

This is one of a little family of slider helpers in the pack (there are fixed 0–1 and 0–100 variants alongside this configurable one); Float Slider is the flexible member, the one you reach for when you want to set the range yourself.

Install comes with the pack. ComfyUI Manager, search SDVN_Comfy_node; or cd ComfyUI/custom_nodes && git clone https://github.com/StableDiffusionVN/SDVN_Comfy_node, then pip install -r custom_nodes/SDVN_Comfy_node/requirements.txt from your ComfyUI root and restart. No weights, no dependencies beyond the pack.

Troubleshooting is basically one thing: the settings JSON has to be valid JSON. A stray comma or a missing brace and the slider falls back or errors, so if your custom range isn't taking, paste the default {"min": 0.0, "max": 1.0, "step": 0.01} back in and edit from there. The other gotcha is a mismatch between the slider's range and what the target expects - a slider capped at 1 can't send a strength of 1.5 no matter how hard you drag, so if a value won't go high enough, it's the max in settings that needs raising, not the node that's broken.