PromptTester

August 13, 2025 ยท View on GitHub

PromptTester Demo

Test and compare LLM prompts across OpenAI, Claude, and Gemini with real-time cost tracking.

Features

  • Multi-Provider: OpenAI (GPT-4/5), Claude (Opus/Sonnet), Gemini (Pro/Flash)
  • Batch Testing: Run multiple scenarios simultaneously
  • Cost Tracking: Real-time pricing with token usage
  • Comparison Mode: Side-by-side response comparison
  • Import/Export: Share scenarios as JSON

Quick Start

# Install dependencies
npm install

# Copy and configure environment variables
cp .env.example .env
# Edit .env and add your API keys

# Start the app
npm start
  1. Configure API keys in .env file
  2. Create test scenarios or use the existing ones in here.
  3. Run and compare results

Configuration

API keys are loaded from environment variables for security. Create a .env file based on .env.example:

REACT_APP_OPENAI_API_KEY=your-openai-key
REACT_APP_CLAUDE_API_KEY=your-claude-key  
REACT_APP_GEMINI_API_KEY=your-gemini-key

How It Works

The app uses official SDKs to connect directly to each provider. It also calculates costs based on each provider's pricing structure.

License

MIT