To put the same image into a ComfyUI batch a specified number of times, add the built-in RepeatImageBatch node, connect your image to image, set amount to the number of copies you want, and use its image output downstream. The node repeats one input image; it does not combine different images or create separate output files by itself.
Decide what the number means
The documentation describes amount as the number of times the input image is replicated. In the RepeatImageBatch implementation at the cited master source, the node applies image.repeat((amount, 1, 1, 1)); for an input batch of N images, this produces N × amount items. Thus, for that implementation, one input image with amount set to 6 yields 6 items: amount multiplies the input batch size rather than adding six items. The source URL points to a mutable branch, not a fixed release or commit, so check the implementation in your installed version when an exact count matters.
This is different from merging distinct images. If you have a portrait and a product photo and want both in one batch, repeating either one will not merge them. It is also different from saving one image repeatedly: a batch is an image output passed between nodes, not a promise that six distinct files will be written.
Build the repeated batch
- In the workflow, place
RepeatImageBatchand connect the single source image to itsimageinput. - Enter the desired copy count in
amount. For a target of 6 identical images, setamountto 6. - Connect the node's
imageoutput to the next node that accepts anIMAGEinput. Keep the connection direct while checking the count, so another operation does not make the result harder to interpret. - If your workflow provides an observable point for the
RepeatImageBatchoutput, check it before connecting the batch to any node that may transform images. Confirm that the output contains N ×amountitems and that each item matches its corresponding input. If the direct output is not observable, use an appropriate downstream node or output that exposes the batch, and treat any differences introduced after an image-transforming node as evidence about that downstream node—not, by themselves, aboutRepeatImageBatch. Then connect the batch to the intended downstream node and check that node's behavior separately.
Check count and identity
Use this worksheet to distinguish the requested batch from other meanings of “repeat.” The figures below are hypothetical examples, not test results. As an author-recommended check, compare the count and image contents at the earliest observable point in your workflow when such a point is available; otherwise use a downstream observation point and record which transformations occurred first. You may compare the result again after downstream processing to locate where a discrepancy first appears, but this is not an official test or a guarantee about custom nodes.
| Input images | amount | Expected output batch | What to verify |
|---|---|---|---|
| 1 source image | 6 | 1 × 6 = 6 | Count is 6; each item shows the same source content |
| 1 source image | 1 | 1 × 1 = 1 | Count is 1; do not expect six items |
| 2 different images | 6 | 2 × 6 = 12 | The two input items are repeated; this does not merge their image content |
A matching count alone is insufficient. If your workflow exposes the RepeatImageBatch output for inspection, compare the copies there, before any image-transforming downstream node. If it does not, use a downstream node or output that makes the batch observable and check both count and contents; however, differences after a transforming node cannot by themselves show that RepeatImageBatch produced different copies. Check the batch count separately, and do not infer batch membership from filenames alone.
If the result does not match
If the batch is too small or too large, compare amount with the target and confirm that the downstream step receives this node's output. If items differ at the direct output, inspect the input and amount; if they differ only after a downstream node, check that node separately before attributing the change to repetition.
If you need several different source images together, use a workflow documented for combining images rather than treating amount as a merge control. If you need separate files, check the save node and workflow: this node's documented output is an IMAGE batch, not a file-count or naming rule. The RepeatImageBatch reference defines its image and amount inputs and its batch output. It does not establish limits for amount, sampler-specific behavior, file naming, or the effects of downstream nodes. The official node documentation also notes that the page is AI-generated, so consult the node behavior in your own workflow when a detail is not specified there.
Common questions
Does repeating an image make separate saved files? Not by itself. The node outputs an image batch; whether and how a later node saves items depends on that node and the workflow.
Can I use this node to combine different pictures? No. Its documented purpose is to replicate its image input. Use an image-combination workflow when the goal is to bring distinct images together.
Should I compare copies after a downstream image operation? For checking the repeated copies, compare the direct RepeatImageBatch output before any operation that may transform them. Downstream results should be evaluated according to that node's behavior.
Sources and next step
- RepeatImageBatch - ComfyUI Built-in Node Documentation
- ComfyUI/comfy_extras/nodes_images.py at master · Comfy-Org/ComfyUI · GitHub
This guide interprets public documentation and uses explicitly hypothetical acceptance examples, not application tests. Controls and availability can change; verify the cited manual against your installed version.
For work created in Imagild, carry the input requirements and acceptance checklist into the photo editor. The external software controls discussed above belong to their respective products.