I Built Flappy Bird Using Every AI Model (You Wont Guess Who Won)

I Built Flappy Bird Using Every AI Model (You Wont Guess Who Won)

---

---

AI Summary

Flappy Bird Benchmarking

Overview

• The video discusses benchmarking various models by recreating the game Flappy Bird.

Models Tested

• Gemini 2
• Gemini 2 Flashing
• 01
• 03
• 03 Mini
• 03 Mini High
• Upcoming models: Sonnet 37, Sonnet 35, Gemini 25, Grok 3.

Testing Process

• Opened terminals for different models, running Python main.py to test performance.
• Various prompts were used across models for consistency.

Performance Ratings

• Gemini 2 Flashing: D
• 01: F
• 03 Mini: D
• 03 Mini High: B
• Upcoming tests with Sonnet 37, Sonnet 35, and others to see performance comparisons.

Observations

• Problems noted included lag, poor interface, and model variations in gravity settings.
• Surprising outcomes with initial models performing better than expected.

Wrap-Up

• Video concludes with thoughts on AI performance in game creation, alongside encouragement for viewers to engage with comments and subscribe for more tutorials.