DeepSeek V3.1 vs Mistral Large 3

Side-by-side benchmark comparison across knowledge, coding, math, and reasoning.

Quick Verdict

Mistral Large 3 wins overall with a score of 61 vs 24 (37 point difference).Mistral Large 3 wins 4 out of 4 categories.

Knowledge

Mistral Large 3

DeepSeek V3.1

30.8

Mistral Large 3

73.8

33
MMLU
76
32
GPQA
75
30
SuperGPQA
73
28
OpenBookQA
71

Coding

Mistral Large 3

DeepSeek V3.1

25

Mistral Large 3

68

25
HumanEval
68

Mathematics

Mistral Large 3

DeepSeek V3.1

32

Mistral Large 3

75

33
AIME 2023
76
35
AIME 2024
78
34
AIME 2025
77
29
HMMT Feb 2023
72
31
HMMT Feb 2024
74
30
HMMT Feb 2025
73
32
BRUMO 2025
75

Reasoning

Mistral Large 3

DeepSeek V3.1

30

Mistral Large 3

72

31
SimpleQA
73
29
MuSR
71

Frequently Asked Questions

Which is better, DeepSeek V3.1 or Mistral Large 3?

Mistral Large 3 scores higher overall with 61 vs 24, a difference of 37 points across all benchmarks.

Which is better for knowledge tasks, DeepSeek V3.1 or Mistral Large 3?

Mistral Large 3 leads in knowledge tasks with an average score of 73.8 vs 30.8.

Which is better for coding, DeepSeek V3.1 or Mistral Large 3?

Mistral Large 3 leads in coding with an average score of 68 vs 25.

Which is better for math, DeepSeek V3.1 or Mistral Large 3?

Mistral Large 3 leads in math with an average score of 75 vs 32.

Which is better for reasoning, DeepSeek V3.1 or Mistral Large 3?

Mistral Large 3 leads in reasoning with an average score of 72 vs 30.