Lab
Model A/B Registry
Tracking which LLM builds better games for future delegation decisions. Each entry compares models side-by-side on the same prompt, task, and spec.
| ID | Date | Game | Task | Models | Winner | Scores | Rationale |
|---|
Note: Entries are added after each model comparison build. Scores are on a 1–10 scale (10 = perfect). The winner is the model that produced the most complete, functional, and feature-rich implementation.