Prompt Optimizer

Live Website

A production-ready, 100% client-side web application for compressing and optimizing LLM prompts using TOON encoding. Reduce token usage by 30-60% while preserving your prompt's core intent - all processing happens in your browser, your data never leaves your device.

Features

• TOON Encoding: Advanced compression using TOON format to reduce token usage by 30-60%
• Multiple Strategies: Choose from Gentle, Balanced, or Aggressive compression modes
• Real-time Token Counting: Live token counting before and after optimization
• 100% Private: All processing happens in your browser - no data ever leaves your device
• Privacy-First: No API calls, no backend, no data collection
• Responsive Design: Mobile-first, works on all devices
• Beautiful UI: Modern dark theme with gradient accents

Getting Started

Prerequisites

• Node.js 18+
• npm or yarn

Installation

# Install dependencies
npm install

# Start development server
npm run dev

Open http://localhost:3000 in your browser.

Build for Production

npm run build
npm start

Project Structure

app/
├── prompt-optimizer/
│   ├── page.tsx       # Main optimizer UI
│   ├── layout.tsx     # Shared layout with nav/footer
│   └── loading.tsx    # Skeleton loader
components/
├── prompt-optimizer/
│   ├── PromptInput.tsx
│   ├── OptimizedOutput.tsx
│   ├── TokenCounter.tsx
│   ├── OptimizationControls.tsx
│   └── CopyButton.tsx
lib/
├── prompt-optimizer/
│   ├── tokenUtils.ts           # js-tiktoken wrappers
│   ├── compressionStrategies.ts # Rule-based + smart compression
│   └── types.ts                # TypeScript interfaces
utils/
├── clipboard.ts               # Safe copy-to-clipboard
└── localStorage.ts            # User preferences

Usage

1. Visit the Live Website
2. Paste your LLM prompt in the input area
3. Select compression strategy:
   • Gentle: Light compression, preserves most meaning (~20-30% reduction)
   • Balanced: Moderate compression, good balance (~40-50% reduction)
   • Aggressive: Maximum compression, significant reduction (~70-80% reduction)
   • TOON: Encode using TOON format for maximum efficiency
4. Click "Optimize Prompt"
5. View the token savings and copy the optimized result

Technical Details

• Token Counting: Uses js-tiktoken with GPT-4o encoding
• Compression: Rule-based heuristics with smart truncation
• Framework: Next.js 14 with App Router, TypeScript, Tailwind CSS
• Browser Support: Chrome, Firefox, Safari, Edge (modern versions)

License

MIT