ComfyUI Node

Searge Output Node

The simplest way to peek at what the LLM wrote

By SeargeDP·Created 2 years ago·Updated 8 months ago· 162
Searge Output Node
  • text

    This is the plainest node in the whole pack, and there's no shame in that - sometimes you just need a place to dump text and look at it. Searge_Output_Node takes one input, of any type, and does nothing with it except display it. No processing, no output socket, nothing downstream. It's a leaf on the graph by design: is_output_node is true and the outputs list is empty.

    What it's for

    Wire the generated output of Searge LLM Node here and you get a quick, dedicated way to read whatever prompt the LLM actually produced, right in the ComfyUI canvas, without hunting through a console log or reaching for a general-purpose text-preview node from some other pack. Since the pack already ships it, it's the path of least resistance if you'd rather not add another dependency just to see a string.

    The one input that matters

    • text - typed *, ComfyUI's wildcard type. That's deliberate: you can plug in a STRING (the obvious case - generated from Searge_LLM_Node), but nothing stops you wiring in anything else the graph happens to be carrying at that point. There's no output socket at all - whatever you send in stops here.

    That's the entire schema. No widgets, no options, nothing to misconfigure.

    Installing it

    Same install as the rest of the pack - this node ships in the same repo as Searge_LLM_Node and Searge_AdvOptionsNode, so there's nothing separate to fetch.

    ComfyUI Manager: search "Searge-LLM for ComfyUI", install, restart. Manually:

    cd ComfyUI/custom_nodes && git clone https://github.com/SeargeDP/ComfyUI_Searge_LLM
    

    Then restart ComfyUI. Unlike Searge_LLM_Node, this one needs no llama-cpp-python, no GGUF model, no models/llm_gguf folder - it's pure display logic, so it works even before you've sorted out the LLM install. If you're setting the pack up for the first time and just want to confirm the custom node loaded at all, this is actually a decent low-stakes node to test with first, before you go chase down the llama-cpp wheel headaches that come with the main node.

    Common issues & troubleshooting

    There's genuinely not much to go wrong here - it's the simplest node in the pack, which is presumably why it barely shows up in search traffic at all. The handful of things worth knowing:

    • You expected it to pass the text onward and nothing downstream is receiving anything. That's correct behavior, not a bug - this node has zero outputs. It's a terminal display, not a pass-through. If you need the text to keep flowing elsewhere (say, to a save-to-file node or a second encoder), tap it off the source - Searge_LLM_Node's generated output - before it reaches this one, and wire that same output to both places.
    • Nothing visibly displays. Make sure the workflow has actually been queued and run to completion; a node that hasn't executed yet has nothing to show. If you're testing in isolation, check that text has a real upstream connection and isn't dangling.
    • You wanted something fancier - syntax highlighting, a copy button, multi-field layout. This node doesn't do any of that; it's intentionally minimal. If you need a more polished text-preview widget, most general-purpose utility packs (rgthree, WAS Node Suite, and others) ship their own "Show Text" node - this one's only real advantage is that it's already installed the moment you have Searge-LLM at all.
    CategorySearge/LLM

    Inputs (1)

    NameTypeDefaultDescription
    text*

    Outputs (0)

    No outputs