Deterministic Inference

Run a deterministic EigenAI inference alongside sampled OpenAI runs to highlight how EigenAI creates a seed-controlled response you can reproduce and verify any time.

Setup

1. Copy .env.example to .env.local and supply provider keys:

   OPENAI_API_KEY=sk-...
   EIGENAI_API_KEY=sk-...
   NEXT_PUBLIC_EIGENAI_DEFAULT_SEED=42
   

   Optional: adjust OPENAI_MAX_TOKENS / EIGENAI_MAX_TOKENS.

2. Install dependencies and start the dev server:

   pnpm install
   pnpm dev

   Visit http://localhost:3000.

Features

• Side-by-side OpenAI sampling vs EigenAI seeded response
• Third EigenAI column generated from a UUID-derived seed to show variation
• Token usage details per run
• Light/dark themes with persistent toggle
• Reset action to clear state after a comparison

Commands

• pnpm dev – run locally
• pnpm lint – lint codebase
• pnpm build / pnpm start – production build & serve

Tech Stack

Next.js App Router, React 19, Tailwind CSS, shadcn-inspired UI components, lucide-react icons.