Kimi K3.1 Failed All My Tests: Opus 5.5 and GPT-6.1 Sol Far Ahead

I tested the new Kimi K3.1 Pro, the second highest rated AI model on Artificial Analysis, with only 5 dollars to build the same three 3D games I gave to Claude Opus 5.5, GPT-6.1 Sol, Gemini 4 Argon and DeepSeek V4.1 Flash. Despite its high benchmark scores, Kimi did not make it to the top 2 in my gaming tests. First, we see what Kimi K3.1 built: a floating mini golf course, a dinner table racing game, and a 3D 8-ball pool game. Then we compare its mini golf with the courses from the other four models, and what each one cost. After that, we race tiny toy cars across a dinner table with all five models. Next, we play the hardest test, a full 3D 8-ball pool game. Then we reveal the final scoreboard, the winners of each game, and where Kimi K3.1 lands. Finally, I share my honest opinion on these results, and the caveats to keep in mind. This is part of my quest to find the best AI model in 2026, tested my own way. Subscribe to the channel, as I cover every update related to the top models, be it from the American or the Chinese AI labs. Timestamps: 0:00 - Kimi K3.1 Pro vs the Top AI Models 1:02 - What Kimi K3.1 Built for 5 Dollars 5:16 - Test 1, 3D Mini Golf Game 8:52 - Test 2, Dinner Table Racing Game 12:21 - Test 3, 3D 8-Ball Pool Game 15:57 - Final Scoreboard and Winners 17:09 - My Honest Opinion #KimiK31 #Kimi #MoonshotAI #ChineseAI #Opus55 #ClaudeAI #GPT61Sol #OpenAI #Gemini4Argon #DeepSeek #AIChallenge #AIGames #AINews #ArtificialIntelligence