Melo AI is an AI-powered music generation application that enables users to create personalized songs through a guided, conversational workflow. Users define mood, genre, and theme, and the platform generates complete music tracks using advanced AI models. The app is designed for scalability, performance, and long-term maintainability using a clean MVVM architecture and a cloud-native backend.
Melo AI provides an end-to-end music creation experience, from onboarding to song generation, playback, discovery, and subscription-based premium features. The product is built with a strong separation of concerns, ensuring predictable state management, testability, and rapid feature iteration.
Users are taken through a structured onboarding flow that captures creative intent:
- Mood selection (e.g. Happy, Chill, Motivational)
- Genre selection (e.g. Pop, Rock, Rap, K-Pop)
- Song theme selection (e.g. Love, Future Self, Pet)
These inputs are compiled into a structured prompt for AI music generation.
- AI-generated vocals and instrumentals
- Instrumental-only generation support
- Multiple generation models (e.g. Melo 3.5)
- High-quality audio output (MP3 and WAV)
- In-app audio player with waveform visualization
- Seek, replay, and skip controls
- Instant preview after generation
- Unlimited song generation
- Voice cloning for personalized vocals
- Priority processing and faster generation
- High-quality audio downloads
- Advanced customization options
- Curated playlists by mood and genre
- Search by song, playlist, or style
- New releases and featured AI-generated albums
- Personal library for saved creations
1. Welcome and onboarding
-
Mood selection
4. Song theme selection
5. AI song generation
6. Subscription upgrade (optional)
7. Explore and library access
This flow minimizes friction for new users while progressively introducing advanced functionality.
Melo AI follows a strict Model–View–ViewModel (MVVM) architecture to ensure a clean separation of concerns between UI, business logic, and data layers.
- UI remains declarative and stateless
- ViewModels handle business logic and state
- Models represent domain and data structures
- External services are abstracted behind repositories
┌──────────────────────────┐
│ View │
│ UI Screens and Widgets │
│ │
│ - Onboarding │
│ - Song Creation │
│ - Player │
│ - Explore / Library │
└────────────▲─────────────┘
│ UI State
┌────────────┴─────────────┐
│ ViewModel │
│ │
│ - Business Logic │
│ - User Intent Handling │
│ - State Management │
│ - Validation │
└────────────▲─────────────┘
│ Domain Models
┌────────────┴─────────────┐
│ Model │
│ │
│ - Song Metadata │
│ - Audio References │
│ - User Preferences │
│ - Subscription State │
└────────────▲─────────────┘
│
┌────────────┴─────────────┐
│ Data and Services │
│ │
│ - Supabase Backend │
│ - AI Generation APIs │
│ - Storage and Auth │
│ - Payments │
└──────────────────────────┘
MIT License © 2026 Kennnnnnyyyyyyyyy
See the LICENSE file for full license text.