AI Prompt Engine v3.0 — User Guide
Prompt Generator for Image and Video Models
Table of Contents
- Introduction
- Interface — Getting Started
- Section 01 — Target AI Model
- Section 02 — Scene Template
- Section 03 — Subject / Scene
- Section 04 — Camera & Lens
- Section 05 — Lighting & Atmosphere
- Section 05b — Material & Surface
- Section 06 — Style, Quality & Composition
- Section 07 — Advanced Parameters
- Section 08 — Negative Prompt
- Sections 09–10 — Keyword Chips
- Section 11 — Extra Keywords
- Section 12 — Output & Smart Copy
- Veo 3 — Video Mode
- Action Buttons
- Workflow — Step by Step
- Example Prompts
- Glossary
1. Introduction
AI Prompt Engine is an advanced tool for building prompts for all major generative AI models. The generator runs locally in your browser — no login required, no data is sent externally.
Supported Models
| Model | Type | Description |
|---|---|---|
| Midjourney | Image | Tag-based with flags like --ar, --v, --style raw |
| DALL-E 3 | Image | Natural language only, no flags |
| Stable Diffusion | Image | Tags + Steps / CFG / Sampler |
| Flux Dev | Image | Descriptive prompt, no MJ flags |
| Adobe Firefly | Image | Natural language optimized for Adobe ecosystem |
| Nano Banana 2 | Image | Verb-first structure, positive framing |
| Veo 3 | Video | Director-style paragraph prompt |
The generator collects parameters from all sections and builds prompts automatically. Image models → comma-separated tags. Veo 3 → cinematic paragraph.
2. Interface — Getting Started
The generator appears at the top of the blog (hero section). Articles are displayed below.
Terminal Window
The interface resembles a system terminal with dark UI. Header includes file name prompt_engine.exe — v3.0 and active model badge.
Compile Animation
// parsing subject... // resolving camera matrix... // injecting lighting rig... // compositing materials... // building prompt architecture... // finalizing output...
The prompt appears after ~0.6 seconds.
3. Section 01 — Target AI Model
This is the most important decision.
- Midjourney — enables MJ parameters
- DALL-E 3 — pure natural language
- Stable Diffusion — CFG, Steps, Sampler
- Flux Dev — descriptive mode
- Adobe Firefly — Adobe-style prompts
- Nano Banana 2 — verb-first mode
- Veo 3 — full video mode
4. Section 02 — Scene Template
Scene Templates automatically fill all fields.
| Template | Description |
|---|---|
| Cinematic | Night street, neon, rain |
| Portrait | Studio portrait, Rembrandt lighting |
| Landscape | Mountain sunrise |
| Macro | Dewdrop close-up |
| Street | Urban life scene |
| Noir | Black & white detective scene |
| Fantasy | Dragon on cliff |
| Futuristic | Sci-fi interiors |
| Product | Studio commercial shot |
You can edit all fields after applying a template.
5. Section 03 — Subject / Scene
The most critical field.
Write in English and be specific.
Good:
elderly Japanese woman in a dimly lit rainy café
Bad:
nice photo
6. Section 04 — Camera & Lens
Camera Examples
- Sony A7R V — ultra sharp
- Hasselblad X2D — luxury look
- Leica M11 — street style
- iPhone — mobile aesthetic
Focal Length
| Lens | Effect |
|---|---|
| 24mm | Wide |
| 50mm | Natural |
| 85mm | Portrait |
| 200mm | Telephoto |
7. Section 05 — Lighting & Atmosphere
Lighting Types
- Golden Hour
- Blue Hour
- Rembrandt
- Neon Glow
- Chiaroscuro
Mood
Misty, Rain, Snow, Stormy, Serene
8. Section 05b — Material & Surface
- Metal (Steel, Aluminium)
- Organic (Wood, Leather)
- Futuristic (Liquid Metal, Carbon Fibre)
Surface: Matte → Glossy
9. Section 06 — Style, Quality & Composition
Styles
- Photorealistic
- Cinematic
- Oil Painting
- Concept Art
Quality
- 8K Masterpiece
- 4K Professional
- RAW
10. Section 07 — Advanced Parameters
Midjourney
--ar 16:9 --v 6.1 --style raw --q 2
Stable Diffusion
- Steps
- CFG Scale
- Sampler
11. Section 08 — Negative Prompt
blurry, low quality, watermark, text, deformed
12. Sections 09–10 — Keyword Chips
Clickable keywords for quick insertion.
13. Section 11 — Extra Keywords
Add custom keywords manually.
14. Section 12 — Output & Smart Copy
- Generate → creates prompt
- Edit → manual modification
- Copy → copy prompt
- Smart Copy → optimized per model
15. Veo 3 — Video Mode
Generates cinematic paragraph prompts:
Medium shot, 35mm lens frames a detective... Lighting: neon reflections... Camera: slow dolly-in...
16. Action Buttons
| Button | Function |
|---|---|
| Generate | Create prompt |
| Edit | Edit manually |
| Copy | Copy output |
| Reset | Clear all |
17. Workflow
- Select model
- Choose template
- Edit subject
- Generate
- Copy
18. Example Prompt
young man, cinematic portrait, 85mm lens, f/1.8, Rembrandt lighting, warm tones, ultra realistic --ar 4:5 --v 6.1
19. Glossary
| Term | Meaning |
|---|---|
| Bokeh | Blurred background |
| Aspect Ratio | Image proportions |
| DOF | Depth of field |
| Prompt | Instruction for AI |
AI Prompt Engine v3.0 — Complete Prompt Generator System