Create consistent AI characters from your photos
NanaTwin is a web application that leverages Google's Imagen 4 (nano-banana) to generate consistent character images across different scenarios. Built with Google AI Studio and Gemini, it provides an intuitive interface for creating professional-looking character shots with precise control over photography parameters.
Watch the complete experience from concept to generation:
https://github.com/user-attachments/assets/overall.mp4
- Character Consistency: Upload up to 15 photos and generate consistent character images across different scenarios
- Custom Trigger Word: Each trained model gets a unique trigger word (e.g.,
dindi) to use in prompts - Photography Controls: Fine-tune your generations with professional parameters:
- 📸 Camera Types: Polaroid/Instant, DSLR, Mirrorless, Film, and more
- 🎭 Photography Styles: Street, Portrait, Documentary, Fashion, etc.
- 🎬 Framing Options: Close-up, Medium shot, Wide angle, etc.
- 😊 Facial Expressions: Neutral, Smiling, Serious, Surprised, and more
- 🎯 Action/Pose: Sitting casually, Standing confidently, Walking, etc.
- 4-Step Workflow: Concept → Upload → Training → Prompting
- Real-time Generation: See your results in a gallery view
Control every aspect of your character generation with intuitive dropdowns
Consistent character across multiple scenarios and settings
- Node.js (v16 or higher)
- A Google Gemini API key (Get one here)
-
Clone the repository
git clone https://github.com/dindicoelho/nanatwin.git cd nanatwin -
Install dependencies
npm install
-
Set up your API key
Create a
.env.localfile in the root directory and add your Gemini API key:GEMINI_API_KEY=your_api_key_here
-
Run the development server
npm run dev
-
Open your browser
Navigate to
http://localhost:3000
- Concept Phase: Define what kind of character you want to create
- Upload Phase: Upload 5-15 photos of your subject (more photos = better consistency)
- Training Phase: Wait for the model to train on your photos (this may take a few minutes)
- Prompting Phase:
- Use your trigger word in prompts (e.g., "A photo of
dindisitting in a cafe") - Select photography parameters from the dropdowns
- Click "Generate" to create your images
- Use your trigger word in prompts (e.g., "A photo of
A photo of dindi sitting in a cafe
Types of Photography: Street photography - Urban, documentary style
Camera: Polaroid/Instant - Nostalgic, unique color rendition
Lens: Standard (35-50mm) - Natural vision, versatile use
Framing: Medium close-up - Head and shoulders visible
Action/Pose: Sitting casually - Relaxed, accessible demeanor
Facial Expression: Neutral/stoic - Minimalist, fashion-forward vibe
- Google AI Studio - Development environment
- Gemini - AI assistance for coding
- Imagen 4 (nano-banana) - Character consistency model
- React/Next.js - Frontend framework
- Tailwind CSS - Styling
I built NanaTwin as an experiment to test Google's Imagen 4 capabilities. While I knew nano-banana was technically superior for character consistency, I was hesitant to use Gemini for coding. This project changed my perspective - the integration between Gemini and Google AI Studio made the development process surprisingly smooth and enjoyable.
Feedback and contributions are welcome! Feel free to:
- Open an issue for bugs or feature requests
- Submit a pull request
- Share your experience using NanaTwin
This project is open source and available under the MIT License.
- Website: View in website
- AI Studio Project: View in Google AI Studio
- Author: @dindicoelho