/TECH

Gemini 3.6 Flash Scores Below 3.5 Flash On Benchmark Leaderboard

BR

Combined views

4.2K

2 posts, first seen 2h ago

73 likes21 comments

Reactions from ranked influencers

2 posts

Combined views

4.2K

2 posts, first seen 2h ago

73 likes comments

4 saves

5 reposts

Sentiment

Positive17.2%82.8%Negative

Positive Read

Many users criticized Gemini 3.6 Flash for regressing below 3.5 Flash on benchmarks, questioning why Google released an inferior model that loses to cheaper competitors.

Based on 12 visible X reactions from 21 accounts; directional sample.

Digg Deeper

Ask a question below.

Published answers will appear here.

Reactions from the X community

S-Shinka - AI@ShinkaIoT

5:51 PM · Jul 21, 2026

@bindureddy Probably just a trial for the real release. I wouldn't surprised if we have a big real release soon. Maybe Gemini 4, or a totally different one. Google isn't staying behind for no reason

AT Alphin Tom@alphinctom

4:52 PM · Jul 21, 2026

@bindureddy Thought i was going crazy with my internal evals..:P Such a horrible update

BI bito@okwhateveritis

5:49 PM · Jul 21, 2026

@bindureddy GOOGLE needs to fire that Demis guy. He's a bad leader.

NƎ NƎO@itz_KGB

4:57 PM · Jul 21, 2026

@bindureddy What tf they did lol, how can they messed up so much 😭🤣

View all

SV Shravan Vadeghar@ShravanVadeghar

5:05 PM · Jul 21, 2026

@bindureddy Even I observed regressions

BR Bindu Reddy@bindureddy

4:22 PM · Jul 21, 2026

1891TECH

🚨GEMINI 3.6 FLASH SCORES BELOW 3.5 FLASH 😅 This is the first time we have seen this in the history of our benchmark! The new version scores below the old version.... So 3.5 appears to be better than 3.6 It also scores below GPT 5.6 Luna, which is cheaper than Flash as well 🤷