AI Prompt Benchmark Tool

This repository contains a minimal prototype of the application described in AGENTS.md. The app lets users submit a prompt, sends it to multiple (dummy) AI models, scores the responses, and displays them side by side.

Features

• Simple web UI for entering a prompt
• Two example models (echo and reverse)
• Length-based scoring metric
• REST API built with FastAPI
• Docker support for easy deployment

Running locally

# Install dependencies
pip install -r requirements.txt

# Start the server
uvicorn app.main:app --reload

Then open http://localhost:8000 in a browser.

Docker

Build and run the Docker container:

docker build -t ai-benchmark-tool .

docker run -p 8000:8000 ai-benchmark-tool

Open http://localhost:8000 to use the app.