Display Text

Display Text does exactly one thing, and the whole reason it ships alongside the MiniCPM VQA nodes in this pack is that those nodes don't show you their answer anywhere on the canvas by default. MiniCPM_VQA and MiniCPM_VQA_Polished both hand back a plain STRING - the model's written answer to whatever you asked it - but a STRING output on its own is just a wire. Without something to render it, you're stuck squinting at logs to find out what the model actually said. Display Text is that something: wire the VQA node's output into it, run the graph, and the answer shows up right on the node itself.

How it works

There's exactly one required input - text (STRING) - and it's marked as an output node, meaning ComfyUI is happy to let it sit at the end of a branch with nothing wired after it. Feed it the STRING from MiniCPM_VQA, MiniCPM_VQA_Polished, or honestly anything else in your graph that produces text, and it renders that string on the node so you can read it directly, no console required.

It also passes the value back out - the output is STRING, list-typed, so it'll happily render more than one entry if whatever's feeding it is itself a list of strings (the MiniCPM VQA nodes aren't - they always hand back one answer per call, however many images went into it). And because it's a passthrough rather than a dead end, you're not forced to choose between "look at the answer" and "keep using it" - wire the output onward to a CLIPTextEncode or a save node and you get both.

Installing it

It's part of the same pack as the MiniCPM VQA nodes, so there's nothing extra to install. ComfyUI Manager: search minicpm, install, restart. Manual:

cd ComfyUI/custom_nodes
git clone https://github.com/IuvenisSapiens/ComfyUI_MiniCPM-V-4_5
pip install -r ComfyUI_MiniCPM-V-4_5/requirements.txt

then restart. This node itself has nothing to do with the vision-language model - no weights, no heavy dependencies - it's just a text display box that happens to live in the same pack. If the pack fails to load entirely (usually a requirements.txt conflict - see the MiniCPM VQA articles for that), this node disappears along with everything else in it, so check your console for an import error before assuming Display Text specifically is broken.

Common issues & troubleshooting

Nothing will connect to it. It only accepts STRING. If you're trying to inspect something else - an INT, a FLOAT, an IMAGE - ComfyUI's type checking blocks the wire outright; you need whatever converts that value to text first.

The box is empty after running. That's almost always upstream, not this node. If MiniCPM_VQA_Polished got an empty or malformed prompt, or the model call failed silently, Display Text is faithfully showing you nothing because there's nothing to show - treat the blank box as a diagnostic pointing back up the chain, not a bug here.

You're seeing several lines instead of one clean answer. The output is list-typed, so if whatever you wired in upstream produces a list of strings rather than one, Display Text renders every item. MiniCPM_VQA and MiniCPM_VQA_Polished each return a single STRING regardless of how many images went into the question, so if you're chaining this after them and seeing multiple entries, look further upstream - something else in the chain is the one handing you a list.