PQ Simple (Batch) / 批量简易

PQ Simple (Batch) is the plain-clothes version of PQ (Batch): same batching logic, preset save/delete removed. It's the node you drop in when you have a list of prompts, want all of them rendered in one pass, and don't need to reorganize your template library mid-workflow.

What it does

Type a list into 多行文本 (one prompt per line), optionally pull it from a .txt via 使用文件 / 提示词文件夹 / 文件名, and it encodes every line and concatenates the conditionings into a single batched positive output. Match the KSampler's batch_size to your line count and each output image gets its own prompt. Set batch_size to 10 for ten lines and you get ten different images - the thing ComfyUI's plain batch count can't do for you.

Prompt assembly is the standard chain: 全局前缀 → 画师风格提示词 → line → 提示词后缀, comma-joined, with the global prefix only applied when you fill it in.

索引设置, batch style

In batch mode there's no counter to advance, so the index field just selects a slice of the list:

• 0 (default) - all lines from the start.
• 5 - everything from the 6th line on (batch indexing is 0-based - different from the incremental node, which counts from 1).
• 2,8,35 - only those lines, also 0-based.

If you're rendering a 200-line file in chunks, this is how you say "just the last 40." Outputs: positive CONDITIONING and prompt_text_list, a newline-joined STRING of what actually went into the batch - handy for a ShowText check before you burn the render.

The one preset touch it keeps is the 预设模板 dropdown, which still reads (but can't write) the templates you saved with the full batch node - Pony, Nova, NoobAI, plus your own. Same built-in list, same {p} placeholder behavior.

Install

No dependencies beyond the torch you already run.

cd ComfyUI/custom_nodes
git clone https://github.com/cyan9977/ComfyUI_PromptQueue.git

Restart, then grab PQ Simple (Batch) / 批量简易 under the PromptQueue category. Parameter labels are Chinese-only; the display name is the bilingual "PQ Simple (Batch) / 批量简易". Two things to keep straight: this is "one pass, many prompts," so if you wanted to loop the same prompt over many seeds, that's the incremental node's job; and if you find yourself wanting to save a template, you'll need the full StylePromptQueueBatch node - this one can only use them.

For a fixed list with no preset bookkeeping, this is the leanest batch option in the pack, and it's the one I'd actually put in a saved workflow.