✦ LaunchKit

From Startup Idea to Live App in Minutes

LaunchKit takes a startup idea in plain text and produces a live, deployed web application. One AI pipeline handles market validation, branding, code generation, and deployment. Powered entirely by Gemini 3.

---

How It Works

1. Describe your idea in plain text
2. Market validation runs competitor analysis and viability scoring
3. Feature brainstorming prioritizes your MVP feature set
4. Brand generation creates names, checks domain availability, builds color palettes
5. Logo design generates custom logos with AI
6. Code generation writes a full React + Tailwind website
7. Live preview renders in a cloud sandbox in real-time
8. Export to Vercel or GitHub with one click
9. Version history tracks every iteration

The entire process takes under five minutes.

---

Tech Stack

Layer	Technology
Frontend	Next.js, React, TypeScript, Tailwind CSS
AI Models	Gemini 3 Pro (validation, features, code), Gemini 3 Flash (naming, colors, logos)
Auth & Data	Firebase Authentication, Firestore
Code Sandbox	E2B Cloud Sandbox
Deployment	Vercel
AI SDK	Google AI SDK (@google/genai)

---

Gemini 3 Integration

LaunchKit uses two Gemini 3 models across the entire pipeline. No other LLM is involved.

Gemini 3 Pro handles:

• Market validation with structured competitor analysis
• Feature brainstorming with MVP prioritization via function calling
• Full website code generation through streaming function calls
• Agentic mode with tool use and self-repair

Gemini 3 Flash handles:

• Brand name generation (sub-second responses)
• Color palette creation based on brand personality
• Logo concept generation

Key Gemini 3 features used:

• Streaming function calls for real-time code generation
• Structured output for market data and feature lists
• Multi-turn tool use in agentic mode
• Automatic error recovery with re-prompting

---

Generation Modes

Fast Mode

Single-shot generation. One API call produces all files. Best for speed.

Agentic Mode

Multi-step generation with tool use. The model plans, writes, tests, and repairs code across multiple turns. Best for complex sites.

Self-Healing Pipeline

If Fast Mode fails, LaunchKit falls back to Agentic Mode. If Agentic Mode encounters errors, it re-prompts with error context. Users see a working result without manual intervention.

---

Project Structure

launchpad/
├── src/
│   ├── app/
│   │   ├── api/
│   │   │   ├── builder/stream/    # Main code generation endpoint
│   │   │   ├── brand/             # Brand name generation
│   │   │   ├── colors/            # Color palette generation
│   │   │   ├── features/          # Feature brainstorming
│   │   │   ├── logo/              # Logo generation
│   │   │   ├── market/            # Market validation
│   │   │   └── export/            # Vercel/GitHub export
│   │   ├── auth/                  # Authentication pages
│   │   ├── builder/               # Builder UI (code editor + preview)
│   │   ├── dashboard/             # Project dashboard
│   │   └── page.tsx               # Landing page
│   ├── components/                # Reusable UI components
│   ├── contexts/                  # React contexts (auth, theme)
│   ├── lib/                       # Utilities (Firebase, Gemini client)
│   └── types/                     # TypeScript type definitions
├── public/                        # Static assets
└── package.json

---

Getting Started

Prerequisites

• Node.js 18+
• npm or yarn
• Google AI API key (Gemini 3)
• Firebase project (for auth and data)
• E2B API key (for code sandbox)

Environment Variables

Create a .env.local file:

# Gemini 3
GOOGLE_AI_API_KEY=your_gemini_api_key

# Firebase
NEXT_PUBLIC_FIREBASE_API_KEY=your_firebase_api_key
NEXT_PUBLIC_FIREBASE_AUTH_DOMAIN=your_project.firebaseapp.com
NEXT_PUBLIC_FIREBASE_PROJECT_ID=your_project_id
NEXT_PUBLIC_FIREBASE_STORAGE_BUCKET=your_project.appspot.com
NEXT_PUBLIC_FIREBASE_MESSAGING_SENDER_ID=your_sender_id
NEXT_PUBLIC_FIREBASE_APP_ID=your_app_id

# E2B Sandbox
E2B_API_KEY=your_e2b_api_key

Run Locally

# Install dependencies
npm install

# Start development server
npm run dev

Open http://localhost:3000 to see LaunchKit.

---

Deploy

Vercel (Recommended)

1. Click the button above or import the repo on Vercel
2. Add environment variables in the Vercel dashboard
3. Deploy

Manual

npm run build
npm start

---

License

MIT

---

Built with Gemini 3 for the Google DeepMind Gemini 3 Hackathon.