Prompt Pack Pixaroma
Paste a list of prompts, queue one run each
- text
You've got a text file with forty prompts and you want one finished image per prompt. Retyping each into a CLIP Text Encode and running forty times is misery. Prompt Pack Pixaroma is the paste-and-forget answer: drop the whole block into the textarea, and it queues one workflow run per prompt automatically. No per-row toggles, no building a loop, no babysitting. This is the node you want when the prompt list already exists somewhere - a document, a spreadsheet column, a brainstorm - and you just need to machine-gun it through your graph.
How it works
A small pill at the top decides how the block splits:
- Paragraph (default) - splits on blank lines. Good for multi-line prompts that are meant to be whole sentences.
- Line - splits on every newline. Good for short, one-line prompt lists.
Each generated image carries the exact prompt that produced it, so if you drag a result back into Prompt Reader Pixaroma you get the right variant back, not a guess. Empty prompts (whitespace only) are silently skipped, and if the textarea is empty when you hit Run nothing queues - you get a warning toast instead of a confusing no-op.
The node has no required inputs or outputs that matter for wiring; the text output simply carries the current prompt for whichever queue run is executing, so you wire it into CLIP Text Encode once and leave it. The real action is in the queue: a counter in the corner shows total prompts, and flips to current/total while a run is in progress so you can see you're on image 7 of 12.
The honest comparison
If you need to toggle individual variants on and off and see them as separate queue items, that's its sibling Prompt Multi Pixaroma (rows, ON/OFF pills, reorder handles). Prompt Pack is the bulk one - you arrive with a finished list and want zero friction. The distinction is worth knowing before you pick: Pack splits a pasted block, Multi holds a hand-curated list. Same family, different job.
Install
It's part of the Pixaroma pack:
cd ComfyUI/custom_nodes
git clone https://github.com/pixaroma/ComfyUI-Pixaroma
or search Pixaroma in ComfyUI Manager, then restart. No extra Python dependencies. Both classic and new node interfaces are supported - though honestly, once you paste your list in and hit Run, you'll mostly be watching the queue panel anyway.
Gotchas
The two real ones: mind the split mode - pasting single-line prompts with Paragraph mode and blank lines in the wrong places will merge rows you meant separate (and vice versa). And remember the output only ever holds one prompt at a time; it's not a list output. If you wire it expecting all prompts at once, you'll get the current run's prompt, which is exactly what CLIP Text Encode wants and nothing else. If the node misbehaves after a pack update, hard-refresh the browser (Ctrl+Shift+R) before blaming it.
Inputs (0)
No inputs
Outputs (1)
| Name | Type | Description |
|---|---|---|
| text | STRING | The current prompt for this queue run (one prompt from the block). Wire to CLIP Text Encode. |