Browse all articles on this site by category.
Getting Started Roadmap
If you’re new here, we recommend reading in this order:
- Why z-image-turbo? — Why we chose this model
- Getting Started with ConoHa AI Canvas — Setting up your environment
- Prompt Fundamentals — Learn word order and emphasis syntax through experiments
- Prompt Design Thinking — How to design your own prompts
- Pose Prompt Collection — Practice with examples
Tutorials
| Article | Level | Description |
|---|---|---|
| Getting Started with ConoHa AI Canvas | Beginner | Start AI image generation with just a browser |
| ComfyUI Workflow Download | Beginner–Intermediate | z-image-turbo workflow (negative prompt pre-configured) |
| RunPod Serverless Setup Guide | Advanced | Build an API environment from scratch |
Reviews & Comparisons
| Article | Description |
|---|---|
| Why z-image-turbo? | Model lineage (SD/Flux/z-image), feature comparison, licensing & NSFW support |
| Cloud GPU Comparison | RunPod vs Vast.ai vs ConoHa AI Canvas — pricing & performance |
Tips & Techniques
Fundamentals
| Article | Description |
|---|---|
| Prompt Fundamentals | Word order rules, emphasis syntax, negative prompts (with experiment images) |
| Prompt Design Thinking | From copying to creating — the 5-element decomposition approach |
| Complete Negative Prompt Guide | Category-by-category standard phrases and customization methods |
Hands-On Experiments
| Article | Description |
|---|---|
| Improving a Beach Bikini Prompt in 8 Steps | Adding one element at a time from a minimal prompt |
| Reproducing NB2 Prompts with z-image-turbo | Converting 10 Nano Banana 2 themes to SD-based generation |
| Prompt Refinement: Comparing “Trim” vs “Add” on 10 Themes | Finding unnecessary words and effective additions |
Prompt Improvement Series
Starting from simple prompts and iterating to achieve the intended image. 3 images generated at each step to confirm stability.
| Article | Theme | Key Point |
|---|---|---|
| Café Snapshot | Window-side café snapshot | Succeeded immediately. Long descriptions for amateur feel worked well |
| Retro Izakaya | Retro izakaya (Japanese pub) | Flash, grain, and paper lanterns all reproduced |
| Night Beach | Night beach cinematic | Backlight + moonlight + dress transparency |
| Library Emo | After-school library emo composition | Resolved fluorescent vs sunset lighting conflict |
| Mirror Selfie | Mirror selfie | Hardest theme. 60-70% success rate |
| Rooftop Fashion | Rooftop fashion magazine style | Solved the “magazine” keyword causing logo artifacts |
| Tokyo Neon Street | Rainy Tokyo neon street | wet pavement reflections was the key |
| Morning Bed Intimate | Morning bed intimate portrait | Soft morning light through curtains with white sheets |
| Summer Festival Polaroid | Summer festival polaroid style | cotton candy in hand stabilized hand rendering |
Body Expression Experiments
| Article | Description |
|---|---|
| The Ultimate Breast Size Prompt Guide | Conclusions from 30+ conditions. Max 11 words, efficient 5 words, plus what doesn’t work |
| Maximizing Breast Size: A Comprehensive Prompt Comparison | Summary of 30+ conditions tested. Best descriptors and key findings |
| Maximizing Butt Size: A Comprehensive Prompt Comparison | 27 conditions tested at front/rear angles. Stacking and body modifier effects |
| Prompts for “Saggy Breasts” | 31 conditions, 186 images. Only size words + spatial descriptions partially effective |
| Prompts for “Flat Chest” | 35 conditions, 192 images. None worked. Breast size reduction not possible with z-image-turbo |
Effective Descriptors
| Article | Description |
|---|---|
| Size Adjective Step-by-Step Comparison | large → huge → gigantic effect. Plateaus at huge |
| Stacking Effect and Token Volume Hypothesis | Stacking adjectives adds size. Even repeating the same word works |
| “oversized” is the Strongest Adjective | Achieve near-maximum with just 5 words — best cost-performance |
| Hiding Nipples Through Outfit Changes | Effects of padded bikini top and tight crop top |
Ineffective Descriptors
| Article | Result |
|---|---|
| Cup Size Notation | G-cup, K-cup, double D → No effect |
| cm Notation | 120cm, 150cm → No effect |
| Weight Syntax | (huge breasts:1.6) → No effect |
| Slang & Metaphors | Bombs, watermelons, and cows appeared |
| Japanese Romaji “oppai” | No effect |
| super/strong Prefixes | No effect alone |
| Unified “breasts” Test | Cannot hide nipples (N=13) |
Prompt Element Experiments
| Article | Description |
|---|---|
| CLIP Chunk Splitting Experiment | What happens beyond 75 tokens. Experimentally testing 2nd chunk influence |
| Does “coherent anatomy” Work? | Tested with 24 images → No effect on z-image-turbo. Recommend removing (saves 7 tokens) |
| Does “actress” Change Faces? Profession Comparison | 7 professions × 3 images. actress/model tend to produce more refined faces |
| God Prompt Ablation Study | Removing elements one by one to verify necessity. Up to 50% token reduction achieved |
Curated Prompts
| Article | Description |
|---|---|
| 3 “God Prompts” That Never Miss | Ablation-verified minimal versions. Café 42 tokens, Summer Festival 48 tokens, Morning Bed 79 tokens |
Prompt Collections
| Article | Images | NSFW | Description |
|---|---|---|---|
| Portrait Prompt Collection | — | SFW | Basic examples by situation |
| Pose Prompt Collection | 12 | SFW | Arms crossed, reading, walking, yoga, etc. |
| Lighting Prompt Collection | 8 | SFW | Same subject, different lighting comparison |
| Expression Prompt Collection | 8 | SFW | Smile, serious, surprised — 8 patterns |
| Gravure Prompt Collection | 12 | NSFW | Swimsuit, lingerie, body line |
Paper Explanations
| Article | Paper | Relevance to AI Image Generation |
|---|---|---|
| CLIP Paper Explained | Radford et al., 2021 | Prompt vectorization, origin of the 75-token limit |
| LDM Paper Explained | Rombach et al., 2022 | Stable Diffusion architecture |
| CFG Paper Explained | Ho & Salimans, 2022 | Theoretical basis of negative prompts |