Box

@box · Twitter · 2026-08-13 17:14 UTC

We put Gemini 3.7 Flash through the Box Complex Work Eval — real, multi-document enterprise tasks graded on whether the finished work is correct. It scored 67% vs Gemini 3.6 Flash's 62%, with the biggest gains on due diligence and data analysis, and it ran about 37% faster. More accurate and lighter to run.

Google DeepMind

@googledeepmind · Twitter · 2026-08-13 17:04 UTC

Gemini 3.7 Flash is here. It’s stronger for coding, knowledge work, and web development. 🧵