Google DeepMind launches Gemini 3.6 Flash with better quality and lower token usage

GoogleDeepMind · x · 2026-07-22

Gemini 3.6 Flash is built directly on feedback from 3.5 Flash, according to Google DeepMind.

The launch claims better quality with lower token usage, and the accompanying chart shows gains across agentic benchmarks such as long-horizon software engineering, machine learning engineering, knowledge work, and computer use.

Related event: Google Launches Gemini 3.6 Flash and Other New Models(65 posts)→

Original post →

More from Models

Models channel →