AI Video Production // Prompt Framework

PROMPT
BUILDER

Tap options to build your prompt. Use text fields for custom details. Hit Generate when ready.

🔑
Anthropic API Key
Required for Image Analyzer and Brief & Fill. Stored locally in your browser only.
✦ Brief & Reference Describe your idea and/or upload a reference image — chips will auto-fill
Your Creative Brief
Reference Image (optional)
↑
Click to upload
Person · Product · Scene · Style ref
Content: Polished, aspirational, brand-driven
01
Subject
Who or what is in frame
Subject Type
Man
Woman
Child
Elderly
Couple
Group
Crowd
Animal
Product
Vehicle
Environment only
Action / Pose
Standing
Walking
Running
Sitting
Lying down
Jumping
Reaching
Looking away
Looking at camera
Mid-action
In conversation
Emotion
Happy
Sad
Angry
Fearful
Surprised
Neutral
Determined
Exhausted
Peaceful
Anxious
Confident
Vulnerable
Stoic
Wardrobe Style
Casual
Formal
Sportswear
Luxury
Streetwear
Uniform
Futuristic
Historical
Gym wear
Beachwear
Night out
Loungewear
Workwear
Outdoor / hiking
Date night
Festival / boho
Minimal basics
No branding
👤 Character Details ▼
Specific physical details. The more precise you are here the less the AI defaults to generic looks.
Ethnicity / Heritage
White / Caucasian
Black / African
East Asian
South Asian
Southeast Asian
Latino / Hispanic
Middle Eastern
Mixed heritage
Indigenous
Skin Tone
Very fair
Light
Medium / olive
Medium-dark
Dark
Deep dark
Age Range
Teen 16–19
Early 20s
Mid 20s
Late 20s
Early 30s
Mid 30s
Late 30s
40s
50s
60s+
Build / Body Type
Slim / lean
Athletic / toned
Muscular
Average
Curvy
Plus size
Petite
Tall / lean
Attractiveness / Look
Model / striking
Conventionally attractive
Girl / guy next door
Natural beauty
Editorial / unconventional
Approachable / everyday
Rugged / weathered
Clean-cut / polished
Facial Features
Strong jawline
Soft features
Defined cheekbones
Freckles
Dimples
Full lips
Sharp eyes
Wide eyes
Beard / stubble
Clean shaven
Glasses
Tattoos
Hair Color
Black
Dark brown
Medium brown
Light brown
Blonde
Platinum blonde
Red
Grey / silver
White
Dyed vibrant
Hair Style & Texture
Short
Medium length
Long
Straight
Wavy
Curly
Tight coils
Afro
Locs
Braids
Ponytail
Bun
Shaved / buzzcut
Messy / undone
02
Environment
Where & when
Setting
Home interior
Office
Luxury interior
Industrial
City street
Rooftop
Forest
Open field
Mountain
Beach
Desert
Studio
Futuristic
Time of Day
Early morning
Golden hour
Midday
Overcast
Blue hour
Night
Pre-dawn
Weather / Atmosphere
Clear
Rain
Fog
Storm
Snow
Dust
Smoke
Fire
Wind
03
Shot & Lighting
Shot Type
ECU
Close-up
Med close-up
Medium
Med wide
Wide
Extreme wide
Low angle
High angle
Bird's eye
OTS
POV
Depth of Field
Shallow
Deep focus
Foreground blur
Lighting
Natural
Flat / diffused
Hard directional
Soft box
Rim lighting
Backlit
Silhouette
Chiaroscuro
Neon
Practical
Lens flare
Dappled
Aspect Ratio
16:9 Commercial
9:16 Social
1:1 Square
4:5 Portrait
2.39:1 Scope
04
Mood & Style
The soul
Color Grade
Natural
Warm
Cool
Desaturated
Vivid
Teal & orange
High contrast
Low contrast
B&W
Monochromatic
Film Texture
Clean digital
Subtle grain
Heavy grain
Overexposed
Underexposed
Vintage
UGC / smartphone
Hyperrealistic
Director / Cinematographer Reference
Roger Deakins
Lubezki
Hoytema
Villeneuve
Fincher
Wes Anderson
Nolan
Wong Kar-wai
Michael Mann
Malick
Commercial / Ad Style
High-end commercial
Documentary
Fashion editorial
Lifestyle
Product / studio
UGC ad
Luxury brand
Tech brand
Social native
Mood Keywords
Epic
Intimate
Melancholic
Hopeful
Tense
Peaceful
Mysterious
Romantic
Gritty
Raw
Dreamlike
Powerful
Playful
Nostalgic
Unsettling
Euphoric
Lonely
Urgent
🔊 Sound Design Veo generates audio natively — prompt it or it guesses ▼
Veo 3.1 generates high-fidelity audio synced to the physics of the video. If you don't prompt it, the model guesses — or generates silence. Describe what should be heard.
Ambient Profile
Absolute silence
Quiet room tone
Office ambience
Heavy city traffic
Gentle city street
Coffee shop
Forest / nature
Ocean / beach
Wind howling
Rain on glass
Heavy storm
Club / venue
Gym
Late night quiet
Foley & Sync Sound
Footsteps — concrete
Footsteps — soft surface
Glass breaking
Liquid splashing
Fabric rustling
Door open / close
Keyboard typing
Cup on table
Paper rustling
Car engine start
Metal impact
Zipper
Phone notification
Camera shutter
Vocalizations & Non-Dialogue
Heavy breathing
Calm breathing
Distant crowd cheering
Muffled laughter
Child laughing
No voices
Background singing
Crowd murmur
Custom Sound Notes
📱 POV & Camera Perspective ▼
How the camera relates to the subject — especially useful for UGC content.
Selfie / POV Style
Selfie — arm extended
Phone visible in frame
First person POV
Mirror selfie
Phone propped up
Hidden camera feel
Screen recording
Friend filming
Camera Device Feel
iPhone front camera
iPhone rear camera
Android camera
GoPro / action cam
Old phone / low res
Security camera
Subject Awareness
Looking at camera
Glancing at camera
Candid / unaware
Talking to camera
Reacting to camera
🎥 Camera & Equipment Optical character of the image ▼
The camera body, lens, and format shape the entire optical DNA of an image. This is where your directing knowledge separates your prompts from everyone else's.
Camera Body
ARRI Alexa
RED
Sony Venice
Blackmagic
Canon Cinema
iPhone
GoPro
16mm film camera
Super 8 film camera
DSLR
Focal Length
14mm ultra wide
24mm wide
35mm
50mm standard
85mm portrait
105mm
135mm
200mm telephoto
Macro
Lens Type & Character
Anamorphic
Spherical
Vintage glass
Modern clinical
Tilt-shift
Fisheye
Soft focus
Chromatic aberration
Lens breathing
Vignette
Film Format
Full frame
Super 35
IMAX
16mm
Super 8
35mm film
Medium format
Bokeh Character
Oval / anamorphic
Circular / clean
Swirly / vintage
Creamy / smooth
Harsh / busy
Minimal bokeh
Shutter & Motion Feel
180° cinematic
High shutter / sharp
Low shutter / blur
Slow motion
Undercranked
Exposure Style
Correctly exposed
Overexposed / airy
Underexposed / moody
High key
Low key
Silhouette exposed
Filter / In-Camera Effect
Pro Mist diffusion
Black Pro Mist
Polarizer
Star filter
Infrared
No filter / clean
🚫 Negative Prompt — What NOT to generate ▼
Nano Banana 2 note
Nano Banana 2 understands what you do want well enough that negative prompts are often unnecessary. Use them as a safety net — but don't rely on them. Detailed Subject + Camera sections prevent most artifacts on their own.
Text on screen
Watermarks
Logos
Blurry faces
Extra limbs
Distorted hands
Lens distortion
Motion blur
Overexposure
Underexposure
Artificial skin
Cartoon style
Anime style
Low quality
Noise / grain
Multiple subjects
Cropped subject
Cluttered background
🖼️ Reference Image Instructions ▼
Use these when providing a reference image — face, character, product or scene you want preserved.
Face / Identity
Preserve face exactly
Keep identity
Match likeness
Do not change face
Consistent face
Body / Physique
Keep body type
Same physique
Same proportions
Hair
Same hairstyle
Match hair color
Same hair length
Clothing / Wardrobe
Same clothing
Preserve outfit
Match accessories
Product / Object
Preserve product shape
Match product colors
Keep branding
Same scale
Scene / Environment
Same location
Same lighting setup
Match color grade
Same composition
Generated Prompt
Copied!
Tap options above and hit Generate →
Director's Notes
→Style Reference is your most powerful field. One good director reference does more than 10 technical details.
→For Veo: Shot Type + Camera Movement together. That combo separates cinematic from generic.
→UGC formula: Handheld shaky + UGC texture + Overexposed. Fools the algorithm every time.
→Negative prompts: safety net for Veo, mostly optional for Nano Banana 2. Strong Subject + Camera sections prevent most artifacts.
→Workflow: Nano Banana still first → use as image reference in Veo for visual consistency.
Last 20 prompts
Auto-saved when you hit + History
No history yet. Generate a prompt and hit + History to save it.
Saved favorites
Prompts you starred — no limit
No favorites yet. Generate a prompt and hit ★ Save Favorite.
Prompt Analyzer
Paste any prompt — see what's working, what's missing, and why
Paste a prompt above and hit Analyze →
Image Analyzer

REVERSE
ENGINEER
ANY IMAGE

Drop any image — film still, commercial, reference ad, your own output. Get a full cinematographic breakdown and send it straight to the builder.

🎞
Drop an image to analyze
Film stills · Commercials · Ads · References · Your own outputs