🔱 Artha Gemini Response

Artha Gemini Response is the pack's "show me the answer" node. Every other Gemini node returns a string, and a string alone is invisible in ComfyUI - wire it into nothing and you have no idea what the model said. Response is the terminal that displays it. It's an output node, meaning it's meant to sit at the end of a chain and render the text in the UI, and its description is unapologetically direct: its objective is to display other Gemini nodes' outputs.

It has one small superpower over a plain text preview: it cleans the response. Gemini has a habit of larding its output with markdown asterisks and hash marks, and this node strips both before showing you the result. If you're tired of reading **bold** and ## headers in your prompt outputs, this is the node.

What it does

The schema is minimal:

• response - a STRING input. This is the text to display. It's a forced input, so wire in whatever Gemini output you want to see.
• text - a multiline STRING input, blank by default. Think of it as a scratchpad companion; you can use it to annotate or compare.
• response - a STRING output with the tooltip "Cleaned Gemini response." It passes the text through with all * and # characters removed.

So it's a pass-through with a cleanup pass and a display: the cleaned string is shown in the UI and comes out the output, which means you can put it mid-pipeline as a "inspect this then keep going" checkpoint. That's the genuinely useful bit - most display nodes end the chain, this one doesn't have to.

Where it fits

You'll reach for it in the obvious spots: after Artha Gemini Portrait or Compose to read the final prompt before it hits your encoder; after Artha Gemini Prompter to see the enriched result; after any vision or question node whose answer you want to actually read. It's also the natural debugging node - wire the traits output of Portrait or Compose into it to see exactly what attribute list went into the prompt.

It pairs with Artha Gemini Markdown, its sibling display node. Markdown renders the text as formatted markdown; Response strips the markdown characters and shows clean plain text. Response is the one you want when the output is destined for a prompt encoder, where stray asterisks and hashes would pollute the prompt. Markdown is for when you want to read a formatted reference document.

The only trap

It's a viewer and a cleaner, not a fixer. If the upstream node returned an empty string - the pack's signature failure mode, where API errors are printed to the console and an empty string is handed back - Response faithfully cleans and displays an empty box. Nothing throws, nothing tells you what went wrong. An empty display is the clue to go check the ComfyUI console.

And there's no API key involved. This node is pure front-end display; it never calls Gemini, so it's safe to use even before you set up your key.

Install

Comes with the whole 🔱 pack:

cd ComfyUI/custom_nodes
git clone https://github.com/Cyrostar/ComfyUI-Artha-Nodes
cd ComfyUI-Artha-Nodes
pip install -r requirements.txt

Restart ComfyUI, or install "Artha" via ComfyUI Manager. No key needed - this one's free.