Gemini 3.6 Flash will only matter if it is extremely efficient

Angaisb_ · x · 2026-07-21

The author says the key question for Gemini 3.6 Flash will be how efficient it is in practice.

If it is not extremely efficient, they argue that a lower reasoning-effort mode on GPT-5.6 Sol could still be the better choice.

Related event: Gemini 3.6 Flash Shows Slight Token Efficiency Improvement(2 posts)→

Original post →

More from Models

Models channel →