OpenAI's efficiency advantage over Gemini explained

Theo - t3.gg discusses how OpenAI's models score higher on coding benchmarks using a fraction of the tokens compared to Gemini's 250k per task average. He references the 'Grug speak' theory of model efficiency and leaked GPT 5.5 reasoning traces.