Nodes/ComfyUI-ReplenishNodes/Load Sampler Name
ComfyUI Node

Load Sampler Name

The sampler dropdown, minus the model

By hben35096·Created 2 years ago·Updated 12 months ago· 7
Load Sampler Name
    • STRING
    sampler_name

    This node does one thing, honestly: it shows you the exact same sampler dropdown as ComfyUI's KSampler, and when you pick one, it hands the name back as plain text instead of running anything. No model attached, nothing gets sampled. Out of context that sounds pointless - it's because this node isn't meant to sample anything. It's meant to feed other nodes that want a sampler name as a string.

    Why you'd actually want this

    ReplenishNodes is upfront in its own README that most of its nodes are "secondary" helpers for building XY Plot comparison grids alongside kenjiqq's qq-nodes-comfyui (plus images-grid-comfy-plugin and Impact Pack for the heavier lifting). A grid comparing euler against dpmpp_2m against uni_pc across a row of test renders needs each axis value to be exactly right. Typing dpmpp_2m_sde_gpu by hand into a text box is exactly how you get a silent typo and a grid where one column quietly fails or falls back to something else. This node sidesteps that: you pick from the real dropdown, so whatever comes out is guaranteed to be a sampler ComfyUI actually recognizes.

    It's also handy outside grid-building. A common pattern is stamping the sampler name into your output filename or a text overlay, so that months later, staring at a folder of renders, you know which used ancestral sampling and which used something deterministic - without digging through embedded workflow metadata.

    How it works

    The sampler_name widget is populated from ComfyUI's built-in sampler list - the same roughly 44 options (euler, euler_ancestral, dpm_2, dpmpp_2m, uni_pc, and so on) that fill the KSampler's own dropdown. Selecting one doesn't touch a model; the node's entire job is converting your selection into a STRING. That's the only output, tooltip straight from the author: "The sampler name."

    What to set, what it outputs

    One field: sampler_name, a dropdown with around 44 choices. Pick whichever sampler you want represented downstream.

    The output is a single STRING. Wire it into a grid axis/label node, a filename template, a text-concatenation node, or anything else expecting a string rather than a live sampler object.

    Installing it

    Through ComfyUI Manager, search ComfyUI-ReplenishNodes and install. Or by hand:

    cd ComfyUI/custom_nodes
    git clone https://github.com/hben35096/ComfyUI-ReplenishNodes
    

    Restart ComfyUI. The README doesn't call out any Python dependencies or model downloads for this node - it's reading ComfyUI's own built-in sampler list, so there's nothing extra to fetch.

    Where people get tripped up

    The obvious one: this node's dropdown looks identical to a real KSampler's, so it's easy to assume it plugs straight into one. It doesn't - the output type is STRING, and ComfyUI's socket typing will refuse to connect it to anything expecting a live SAMPLER. If you want a graph that actually swaps which sampler runs at execution time, you need something that turns a string back into a sampler object; this node only produces the label, on purpose.

    The other thing worth knowing before you build a grid around it: this pack barely documents itself past the XY-Plot use case in its README. If you're trying to do something this node doesn't obviously support, there's no wiki to check - just the source and whatever you can infer from the widget itself.

    CategoryReplenish/QQNodes

    Inputs (1)

    NameTypeDefaultDescription
    sampler_nameCOMBOThe name of the sampler.

    Outputs (1)

    NameTypeDescription
    STRINGSTRINGThe sampler name.