Skip to content

Pit different AI models against each other in debates. You can assign personalities, control how spicy the arguments get, and save the good ones. Handles multiple LLM APIs with real-time streaming.

Notifications You must be signed in to change notification settings

joshuaisaact/AIgument

Repository files navigation

AIgument

Watch AI models debate! Instead of dry benchmarks, watch two LLMs argue it out and decide who wins.

AIgument Preview

Features

  • Watch different AI models debate each other
  • Choose from various models (e.g., GPT-4o, Claude 3 Sonnet, Gemini 1.5 Flash, Grok 3, Grok 3 Mini)
  • Assign unique personalities to each debater (e.g., Pirate, Butler, Sassy Drag Queen, etc.)
  • Control the "spiciness" or intensity of the debate
  • Vote on which argument you think is better
  • Browse and view previously saved debates
  • Simple, clean interface focused on the debate

Getting Started

  1. Clone the repository
  2. Install dependencies:
    pnpm install
  3. Create a .env.local file with your API keys:
    NEXT_PUBLIC_OPENAI_API_KEY=your_openai_key
    NEXT_PUBLIC_GOOGLE_GENERATIVE_AI_API_KEY=your_google_key
    NEXT_PUBLIC_ANTHROPIC_API_KEY=your_anthropic_key
    NEXT_PUBLIC_XAI_API_KEY=your_xai_key
    
  4. Run the development server:
    pnpm dev

How to Use

  1. Enter a debate topic
  2. Select the models for each debater
  3. Assign a personality to each debater
  4. Choose the desired "spiciness" level
  5. Click "Start AIgument" and watch the debate unfold
  6. Use the "Next Round" button to continue
  7. Vote on which argument you think is better
  8. (Optional) Save the debate to view later

Technologies Used

  • Next.js
  • TypeScript
  • Tailwind CSS
  • Vercel AI SDK
  • OpenAI API
  • Google Generative AI API
  • Anthropic API
  • xAI API

About

Pit different AI models against each other in debates. You can assign personalities, control how spicy the arguments get, and save the good ones. Handles multiple LLM APIs with real-time streaming.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages