Mistral Devstral Small 1.1 vs Gemma 4 26B A4B
Compare Mistral Devstral Small 1.1 by Mistral AI against Gemma 4 26B A4B by Google AI, tested across 54 shared challenges.
Our Verdict
Gemma 4 26B A4B
Mistral Devstral Small 1.1Runner-up
No community votes yet. On paper, Gemma 4 26B A4B has the edge — bigger model tier, newer, bigger context window, major provider backing.
Slight edge
API pricing
Cost per 1M tokens
Mistral Devstral Small 1.1
Input
$0.10
1.3× cheaper
Output
$0.30
1.3× cheaper
Gemma 4 26B A4B
Input
$0.13
Output
$0.40
Mistral Devstral Small 1.1 is cheaper on both — 1.3× input, 1.3× output
Writing DNA
Style Comparison
Similarity
52%
Gemma 4 26B A4B uses 15.3x more headings
Mistral Devstral Small 1.1
Gemma 4 26B A4B
50%Vocabulary58%
17wSentence Length18w
0.56Hedging0.31
4.3Bold6.8
5.7Lists3.3
0.00Emoji0.00
0.05Headings0.74
0.34Transitions0.15
Based on 21 + 27 text responses
Research
What we learned reading every model
FAQ
