Choose batch face swap
You need multiple output image files: one identity across several target photos, or several source identities applied to one target as separate generated outputs.
Workflow decision guide
Batch vs multiple face swap: choose by output shape
Batch face swap creates several output images. Multiple face swap maps several replacement identities into selected people inside one group photo. The names sound similar, but the tasks are not interchangeable.
Fast decision
Count outputs first, then count faces
Choose multiple face swap
You need one mapped group-photo result where selected detected people receive different replacement identities.
Do not use batch for a group mapping
Batch mode organizes repeated image outputs. It does not provide the numbered person-to-reference mapping used inside one group photo.
Do not use multiple for a photo set
Multiple face swap starts from one group photo. It is not the workflow for producing a folder of separate target-image results.
Verified workflow facts
Compare the current limits before upload
| Decision | Batch face swap | Multiple face swap |
|---|---|---|
| Primary result | Multiple generated output images | One group-photo result with selected mapped replacements |
| Input pattern | One face reference to multiple target photos, or multiple source faces to one target photo | One group photo plus one replacement reference for each selected detected face |
| Current image count | Up to 20 images per batch | Up to 10 replacement identity references |
| Accepted image types | JPEG, PNG, WebP, AVIF, HEIC, and HEIF | JPEG, PNG, and WebP |
| Upload limits | 30 MB per image and 200 MB combined | 30 MB per image and 120 MB for the complete task |
| Current cost | 2 credits per generated output | 2 credits per selected face |
| Pre-submission control | Review the displayed output count and exact credit total | Review every numbered target-to-reference mapping and selected-face total |
| Settlement | Each requested output contributes to the displayed task total | Credits are charged only after every selected mapping completes successfully |
Open the batch workspace or the multiple face swap workspace only after the expected result matches the row above.
Cost examples
Use the unit that the workflow actually charges
| Example task | Calculation | Current total |
|---|---|---|
| One identity across 8 separate targets | 8 batch outputs x 2 credits | 16 credits |
| 4 selected people in one group photo | 4 selected faces x 2 credits | 8 credits |
| 5 source identities as separate variations of one target | 5 displayed batch outputs x 2 credits | 10 credits |
These examples apply the current documented rate. The exact amount displayed in the live workspace before submission is authoritative. Use the face swap cost calculator to compare a planned task with current package unit values.
Preflight
Make the workflow choice before the quality test
- Write the expected result in one sentence: several files means batch; several mapped people in one photo means multiple.
- Count outputs or selected faces: use that count to estimate 2-credit units.
- Check formats and totals: batch and multiple accept different file sets and combined upload sizes.
- Run the smallest representative test: use two batch outputs or the most difficult two people in a group before scaling.
- Inspect trial image disclosure: both are image workflows, so the trial watermark and paid-unlock rule applies.
For an account with no completed credit purchase, a delivered trial image contains DeepSwapAI.com and the account's 100% complete email address. Every character is visible and the address is not masked, shortened, hashed, or omitted. A completed purchase makes future image exports watermark-free; it does not change an existing trial file.
How this guide was made
Compared from current submission and settlement rules
The DeepSwapAI Product Team compared the live batch modes, mapped group-photo flow, accepted formats, upload limits, pricing units, pre-submission disclosure, settlement behavior, and image-export entitlement. No output-quality ranking or speed benchmark was inferred. Facts were reviewed on July 16, 2026; see the verification methodology and the dedicated batch guide and multiple-face guide.
Choose the workflow that matches the result
Use batch for several image files or multiple face swap for several mapped people inside one group photo.