4 AIs. 4 Games. One Elimination Tournament.

Four AI models. Four games. One elimination tournament. Kimi K3, Fable 5, Opus 5 and GPT-5.6 Sol all get the same tools and the same budget. But not the same number of attempts — cheaper models can afford more iterations. Every round tests a different kind of movement: 🏃 Subway Surfers — running, dodging and obstacles 🛹 Tony Hawk — ramps, tricks and landing 🏙️ Mirror’s Edge — parkour, wallruns and flow 🕸️ Spider-Man — physics-based web swinging Each model gets $50 per round. The final Spider-Man challenge raises the budget to $200 each. Some games worked immediately. Others needed their entire budget just to fix the character. One model built a giant wooden runner, another became obsessed with backflips, and the final round ended much closer than expected. There’s no automatic winner this time — watch both final cities and decide who deserves it. ⏱️ Chapters 00:00 The challenge and equal-budget rules 00:28 Round 1: Subway Surfers — Kimi K3 01:20 Fable 5 01:52 Opus 5 02:35 GPT-5.6 Sol 03:03 The first elimination 03:16 Round 2: Tony Hawk — Kimi K3 04:12 Fable 5 04:39 Opus 5 05:10 The second elimination 05:22 Round 3: Mirror’s Edge — Fable 5 05:59 Opus 5 06:12 A tie changes the tournament 06:26 Final Round: Spider-Man — Fable 5 07:14 Opus 5 08:00 You decide the winner 🤖 Models: Kimi K3 · Fable 5 · Opus 5 · GPT-5.6 Sol 🛠️ Tools: Godot · Blender 💰 Budget: $50 per model per round · $200 for the final Who won the Spider-Man round — Fable 5 or Opus 5? Let me know in the comments. Subscribe for more honest AI coding battles. #AI #AICoding #GameDev #Godot #Blender #VibeCoding #AIvsAI