Mistral Devstral Small 1.1 vs Gemma 3 12B
Compare Mistral Devstral Small 1.1 by Mistral AI against Gemma 3 12B by Google AI, tested across 55 shared challenges.
Our Verdict
Gemma 3 12B
Mistral Devstral Small 1.1Runner-up
No community votes yet. On paper, Gemma 3 12B has the edge — bigger model tier, major provider backing.
Gemma 3 12B is 10x cheaper per token — worth considering if cost matters.
Too close to call
API pricing
Cost per 1M tokens
Mistral Devstral Small 1.1
Input
$0.10
Output
$0.30
Gemma 3 12B
Input
$0.03
3.3× cheaper
Output
$0.03
10× cheaper
Gemma 3 12B is cheaper on both — 3.3× input, 10× output
Writing DNA
Style Comparison
Similarity
79%
Mistral Devstral Small 1.1 uses 1.5x more transitions
Mistral Devstral Small 1.1
Gemma 3 12B
50%Vocabulary52%
17wSentence Length17w
0.56Hedging0.66
4.3Bold6.2
5.7Lists5.4
0.00Emoji0.00
0.05Headings0.06
0.34Transitions0.23
Based on 21 + 23 text responses
Research
What we learned reading every model
FAQ
