Gemini 2.5 Flash Preview (thinking) vs MiniMax M3
Compare Gemini 2.5 Flash Preview (thinking) by Google AI against MiniMax M3 by MiniMax, context windows of 1.0M vs 1.0M, tested across 20 shared challenges.
Our Verdict
MiniMax M3
Gemini 2.5 Flash Preview (thinking)Runner-up
No community votes yet. On paper, MiniMax M3 has the edge — bigger model tier, newer.
Too close to call
API pricing
Cost per 1M tokens
Gemini 2.5 Flash Preview (thinking)
Input
$0.17
1.7× cheaper
Output
$3.50
MiniMax M3
Input
$0.30
Output
$1.20
2.9× cheaper
Gemini 2.5 Flash Preview (thinking) wins input (1.7× cheaper)·MiniMax M3 wins output (2.9× cheaper)
Writing DNA
Style Comparison
Similarity
44%
MiniMax M3 uses 115.0x more emoji
Gemini 2.5 Flash Preview (thinking)
MiniMax M3
53%Vocabulary55%
14wSentence Length24w
0.38Hedging0.52
4.4Bold5.2
4.1Lists3.0
0.00Emoji1.15
0.00Headings1.21
0.24Transitions0.02
Based on 7 + 27 text responses
Research
What we learned reading every model
FAQ
