ComfyUI Node

Sampler Selector

One dropdown for every sampler, wired as a string

By Stibo·Created 5 months ago·Updated 2 months ago· 9
Sampler Selector
    • sampler
    sampler_name

    Every KSampler node has its own sampler dropdown, which means every sampler in your graph has its own copy of that dropdown. Sampler Selector is the pack's answer to that duplication: one node with the full list of ComfyUI samplers, outputting the selected name as a string you wire into as many KSamplers as you like. Change it in one place, and every sampler that reads it changes with it.

    What it is

    • sampler_name - a dropdown populated from ComfyUI's actual sampler list (the same list the core nodes use), so you get everything: euler, euler_ancestral, dpmpp_2m, dpmpp_sde, ddim, uni_pc, and friends.
    • Output: sampler - the selected name as a string, ready to feed a KSampler's sampler_name input.

    That's the whole node. It's a selector, not a processor - its entire job is to be the single source of truth for "which sampler are we using today."

    Why you'd reach for it

    • Multi-sampler workflows. If you run the same sampler across several KSampler stages (base + refine + hires), one selector keeps them in sync. Change it once instead of three dropdowns.
    • Primitive-style graphs. The output is a plain string, so it plays with anything that accepts a sampler name - including nodes that take strings from anywhere, and workflows built around central "settings" areas.
    • Comparing samplers. In an A/B workflow you can even drive a selector from a switch, though honestly the plain dropdown is usually enough.

    This is a convenience node, not a power tool. It doesn't change how samplers behave, it just centralizes the choice. If your graph has one KSampler, it saves you nothing; if it has several, it saves you from hunting for the one you forgot to update.

    Installing it

    Part of the Nifty Nodes pack:

    cd ComfyUI/custom_nodes
    git clone https://github.com/Stibo/comfyui-nifty-nodes
    

    or search "Nifty Nodes" in ComfyUI Manager, then restart. No models or dependencies.

    Gotchas

    Because it draws from the same list as the core nodes, an outdated ComfyUI will limit what's in the dropdown - if you're missing a newer sampler, update ComfyUI, not the node. Also note the output is a string and KSampler's sampler_name input expects the exact name; as long as you wire the selector straight in, that's automatic. And remember the pack itself targets the newer V3 API - update ComfyUI if the node doesn't appear after install.

    Categorynifty/selectors

    Inputs (1)

    NameTypeDefaultDescription
    sampler_nameCOMBOSelect a sampler to pass as a string to a KSampler node.

    Outputs (1)

    NameTypeDescription
    sampler*