AI Image Prompt Generator for Photorealistic AI Photography
Upload a photo or describe a scene — CookedBanana reverse-engineers it into a hyper-detailed, 8-component JSON prompt that Nano Banana understands perfectly. First-try photorealistic results, every time.
No prompt engineering skills required. Free to start.
Writing a photorealistic prompt is harder than it looks
Nano Banana can produce stunning results — but only when you describe exactly what you want. Camera angle, skin texture, fabric grain, light direction, iPhone simulation... most people describe 10% of what the model needs.
Manual prompting
- You describe 10–20% of the parameters
- Inconsistent outputs — different face every time
- No outfit reference system
- 15+ minutes per prompt, multiple retries
- Missing light physics, camera simulation, skin detail
With CookedBanana
- 100% of parameters covered — 8 photographic dimensions
- Subject identity locked across generations
- 14 garment slots with photo references
- Generated in under 10 seconds, photorealistic first try
- iPhone pipeline, light physics, environment simulation included
Three steps to a perfect Nano Banana prompt
Under the hood, CookedBanana runs a highly complex system of data calculation and interpretation — analyzing hundreds of visual and semantic parameters simultaneously to produce ultra-detailed, ultra-realistic prompts. From biometric geometry and fabric microstructure to environmental light physics and lens simulation, every variable is computed and encoded so Nano Banana has everything it needs to generate a perfect shot on the first try.
Step
01
Upload or describe
Drop a photo or type a scene description. CookedBanana accepts both as input — any format, any resolution.
Step
02
Configure references
Lock biometric data to preserve identity. Select outfit fields to reference from additional photos. Toggle the settings you need.
Step
03
Generate & use
Copy the prompt payload and paste it into Nano Banana. Attach reference photos in the listed order. Done — photorealistic output, first try.
8 photographic dimensions in every prompt
Each prompt is a structured JSON with 8 specialized components — covering everything from camera physics to iPhone simulation. Here's what the AI generates for you.
Visual Hierarchy & POV
Camera angle, focal length, resolution, shot type
Subject Identity
Age, ethnicity, posture, face detail, piercings, tattoos
Outfit Full Body
Headwear through footwear — 14 garment slots
Time & Weather
Action, secondary movements, narrative charge
Environment & Clutter
Architecture, props, atmosphere, temperature
Light Physics
Marked light sources, reflected lights, shadow mapping
iPhone Pipeline
Device model, HEIC artifacts, color balance, pro notes
Nano Banana Payload
Final assembled prompt optimized for generation
Everything you need for reverse prompt engineering
Four specialized input modes that cover every image prompting scenario — from photo analysis to identity preservation.
Photo Analysis
Upload any photo and the AI reverse-engineers every detail — subject, outfit, lighting, environment — into a structured JSON prompt with 8 photographic components.
Text to Prompt
Describe the scene in plain text and the AI fabricates all technical parameters automatically. No photography knowledge required.
Lock Ref
Freeze biometric data from your photo. Keep the same face and body while changing outfit, scene, or lighting. Your subject identity stays consistent across every generation.
Outfit References
Tag specific clothing items as references. Nano Banana will clone each piece from the photos you attach — jackets, shoes, accessories, everything.
Your prompt is ready before the render even starts.
Upload any photo. In seconds, CookedBanana writes a prompt so precise it would take you an hour to craft by hand — capturing light, mood, skin texture, every detail your eye caught but your words couldn't find.
By the time Nano Banana finishes rendering, your prompt has been ready for minutes. The bottleneck was never the AI — it was the words. We fixed that.
Try it with your photo
Built for anyone who creates with AI images
AI Photographers
You generate portraits and scenes with Nano Banana but spend more time on the prompt than on the creative direction. CookedBanana writes the technical prompt so you focus on the art.
Content Creators
You need consistent, on-brand AI imagery for social media, ads, or editorial content. Lock your subject identity and generate variations on demand.
Studios & Agencies
Your team needs repeatable, high-quality AI photography across multiple client projects. One tool, consistent prompts, project organization built in.
Simple, transparent plans for AI creators
LLM Prompt Builder included in every plan. Start free, upgrade when ready.
Pro
- 700 generations/month
- LLM Builder included
- Unlimited projects
- JSON export
- Priority support
Agency
- 2,500 generations/month
- LLM Builder included
- Unlimited projects
- JSON export
- Dedicated support
Frequently asked questions
Everything you need to know about the Image Prompt Generator.
It is a specialized AI tool that takes a photo or a text description and generates a hyper-detailed, structured JSON prompt optimized for the Nano Banana image generation model. The generated prompt covers 8 photographic dimensions — from camera angle to iPhone pipeline simulation.
Nano Banana is an AI image generation model that produces photorealistic outputs when given properly structured prompts. CookedBanana's Image Prompt Generator is purpose-built to create the exact prompt format Nano Banana expects.
The prompt structure is optimized specifically for the Nano Banana model. While the textual descriptions are detailed enough to inspire prompts for other models, the JSON structure and payload format is designed for Nano Banana's input requirements.
Lock Ref freezes the biometric identity of your subject — face geometry, skin tone, body proportions. This lets you change the outfit, lighting, or scene while keeping the same person consistent across multiple generations.
You can upload one main photo as input, plus additional reference photos for outfit items. The AI will analyze the main photo and tag the reference garments in the correct output slots.
You can sign up and try the tool for free with a limited number of generations. No credit card required. Upgrade whenever you need more volume.