Batch String ๐
The manual entry point into cgem156's batch-prompt pipeline
- texts
- BATCH_STRING
The simplest way into cgem156's batch_condition family: an expandable list of text entries that collects into a single BATCH_STRING output. Where RandomColorPrompt auto-generates variants and PrefixString transforms an existing batch, BatchString is the plain, manual option - you type each entry yourself, and it becomes one item in the batch.
How it fits the pack
This is typically where a batch-prompt workflow starts. From here, the batch goes into CLIPTextEncodeBatch to encode all entries into conditioning at once, then through a batched sampler, then out through SaveImageBatch and/or SaveBatchString. You can also route it through PrefixString first if you want a shared prefix applied to every hand-typed entry without retyping it into each one.
Inputs and outputs that matter
texts- an autogrow list of string entries. Add as many as you need; each one becomes a separate item in the output batch. This is the pack's variable-length list mechanism, the same oneAttentionCouple'sconds/masksinputs use - add or remove entries in the UI and the node's socket count follows.- Output:
BATCH_STRING- ready forCLIPTextEncodeBatch,PrefixString,SaveBatchString, or anywhere else in the pack that consumes a batch of strings.
Installing it
ComfyUI Manager โ search cgem156-ComfyUI๐ โ install โ restart. Or:
cd ComfyUI/custom_nodes
git clone https://github.com/laksjdjf/cgem156-ComfyUI
then restart ComfyUI. Nothing extra to download for this node - it's plain text handling, no models involved.
Common issues & troubleshooting
Only getting one output when you expected a batch. If you've only added one entry to texts, you get a batch of one - which is valid, but check you actually clicked to add more entries if you meant to build a real batch. The autogrow UI needs an explicit action to add each new slot; it doesn't create extras on its own.
Batch size mismatch with your latent batch. Whatever number of entries you put into texts becomes your conditioning batch size downstream. If your latent_image batch doesn't match, expect shape errors or unexpected broadcasting at the sampler - keep the counts aligned.
Wanting variation instead of manual entries. If you're typing out ten near-identical prompts by hand and it's getting tedious, that's a sign you might want RandomColorPrompt (for color variation specifically) instead - BatchString is deliberately the "you write every word" option, with no generation or randomization built in.
Entries getting reordered or lost. Since this is a UI-managed autogrow list rather than a single text field, be careful when copying workflow JSON between ComfyUI versions or instances - verify the entry count and order survived the copy before running a long batch on it.
Inputs (1)
| Name | Type | Default | Description |
|---|---|---|---|
| texts | COMFY_AUTOGROW_V3 | โ |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| BATCH_STRING | BATCH_STRING | โ |