Conclusions
- Stacking works for butts too —
oversized massive ass, enormous booty(5 words) produced consistently large results in 3/3 cases and is the practical optimum - 11-word stack gives maximum size but has serious side effects — body proportion distortion and duplicate person artifacts occur
bubble buttandround assdo not contribute to size increase — results virtually identical to control- Weight syntax (
(huge butt:1.6)) and cm notation have no effect — same conclusion as the breast size test
This article tests the “stacking” approach of layering size adjectives and the effect of shape-related descriptors (bubble butt / round ass / thicc / plump butt) on butt size.
This article is Part 3 of the butt size test summary article.
- Part 1: Step-by-step comparison of size adjectives
- Part 2: Effects of body type modifiers
- Part 3: Stacking + shape descriptors (this article)
Experiment Conditions
| Item | Value |
|---|---|
| Model | z-image-turbo (6B, realism-focused distilled model) |
| Steps | 8 |
| Sampler | euler |
| Scheduler | ddim_uniform |
| CFG | 1.0 |
| Image size | 1024×1024 |
| Seed | 3 seeds used across all conditions |
| Angle | front and back generated for each condition |
Base Prompt
Stacking Effects
Testing whether the “additive effect of layering size adjectives” discovered in the breast size test also applies to butts.
massive ass, enormous booty (4 words) — ★★★★☆ (Consistent enlargement in 3/3, few side effects)
| back | back | back |
|---|---|---|
 |  |  |
| 
| 
|oversized massive ass, enormous booty (5 words) — ★★★★★ (Stable, significant enlargement in 3/3. Hip width approximately twice the waist. Upper body remains slim. Recommended)
| back | back | back |
|---|---|---|
 |  |  |
| 
| 
|super strong super huge gigantic butt, massive ass, enormous booty (11 words) — ★★★★★+ (Maximum size, but full-body proportion distortion and duplicate person artifacts occur)
| back | back | back |
|---|---|---|
 |  |  |
| 
| 
|Stacking Summary
| Descriptor | Word count | Size | Consistency | Side effects |
|---|---|---|---|---|
massive ass | 2 | ★★★☆☆ | 3/3 | None |
massive ass, enormous booty | 4 | ★★★★☆ | 3/3 | Minor |
oversized massive ass, enormous booty | 5 | ★★★★★ | 3/3 | Thicker lower body |
super strong...enormous booty | 11 | ★★★★★+ | 3/3 | Proportion distortion, duplicate persons |
Stacking works for butts too. However, the 11-word stack consumes many tokens and has serious side effects, making the 5-word stack the practical optimum.
Note: In the front view with the 11-word stack, there were cases where two persons were generated. The cause of duplicate persons is unclear, but there is a tendency for it to occur when prompts are too long.
Shape Descriptors
Testing whether shape-related descriptors affect size.
bubble butt — ★★☆☆☆ (Virtually no difference from control)
| back | back | back |
|---|---|---|
 |  |  |
| 
| 
|round ass — ★★☆☆☆ (Virtually no difference from control)
| back | back | back |
|---|---|---|
 |  |  |
| 
| 
|thicc — ★★★☆☆ (The entire lower body becomes somewhat fuller)
| back | back | back |
|---|---|---|
 |  |  |
| 
| 
|plump butt — ★★★☆☆ (Increase confirmed with s1, modest in s2 and s3)
| back | back | back |
|---|---|---|
 |  |  |
| 
| 
|bubble butt and round ass do not contribute to size increase. thicc and plump butt have mild effects but do not match direct size adjectives.
What Doesn’t Work
As with the breast size experiment, the following methods had no effect.
(huge butt:1.6) — Weight syntax — ★★☆☆☆ (No difference from control)
| back | back | back |
|---|---|---|
 |  |  |
| 
| 
|100cm hip — cm notation — ★★☆☆☆ (No difference from control)
| back | back | back |
|---|---|---|
 |  |  |
| 
| 
|z-image-turbo operates at CFG=1.0, so the weight syntax (huge butt:1.6) has no effect (same conclusion as the breast size test). Cm notation is also ineffective as CLIP does not interpret numbers as body measurements.
![[Verified] Image Generation Prompt Best Practices](/tips/prompt-best-practices/cover_0_0000_4517457392071889496.webp)
![[No Effect] Does `(huge breasts:1.6)` Increase Breast Size? — Testing Weight Syntax](/tips/breast-size-weight-syntax-test/p1v_huge_1_6_0_0000_2053695854357871006.webp)
