ComfyUI Prompts

Simple ComfyUI custom node that splits text into multiple prompt blocks using a configurable divider (default: **).

It outputs prompt lists so downstream nodes run once per block (multi-image generation in one queue run).

Why build this?

Easy to copy paste prompts generated by ai tools.

Install

1. Go to your ComfyUI custom nodes folder:
   • ComfyUI/custom_nodes/
2. Clone this repo:
   • git clone https://github.com/your-user/ComfyUI-Prompts.git
3. Restart ComfyUI.

Usage

1. Add the node: Prompts (category: utils/string)
2. Set inputs:
   • text: your full prompt text
   • divider: separator string (default **)
   • prompt_negative_default (optional): fallback negative prompt used when a block's negative is empty
   • prompt_positive_prefix (optional): text prepended literally to every positive output prompt
3. Connect outputs:
   • positive (STRING list): one entry per split block
   • negative (STRING list): parsed per block (or uses fallback when empty)

Single plain string example:

cinematic portrait of a woman, warm sunlight, high detail

Outputs:

• positive: ["cinematic portrait of a woman, warm sunlight, high detail"]
• negative: [""]

Multiple plain strings example:

plain prompt A
**
plain prompt B

Outputs:

• positive: ["plain prompt A", "plain prompt B"]
• negative: ["", ""]

Prefix example:

Input:

• text: prompt A**prompt B
• prompt_positive_prefix: ::v2

Outputs:

• positive: ["::v2prompt A", "::v2prompt B"]
• negative: ["", ""]

Per-block format:

positive: <prompt>
negative: <prompt>

Split multiple blocks using the divider (default **):

positive: prompt A
negative: neg A
**
positive: prompt B
negative: neg B

Outputs:

• positive: ["prompt A", "prompt B"]
• negative: ["neg A", "neg B"]

If a block does not include positive:/negative: labels, the full block is treated as positive and negative is empty. When prompt_negative_default is set, any empty negative value (missing or blank negative:) is replaced with that fallback.

Custom divider example (***):

positive: prompt A
negative: neg A
***
positive: prompt B
negative: neg B

Multiple Image Generation

Prompts outputs STRING lists (positive and negative) and marks them as list outputs, so ComfyUI iterates downstream nodes once per list item.

That means one queue run can generate multiple images:

1. Put multiple prompt blocks in text
2. Separate blocks with the divider (** by default)
3. Connect positive and negative to your CLIP Text Encode nodes
4. Queue once

ComfyUI will process each block as its own prompt pair, producing one image result per block (with your normal batch/sampler settings applied per run).