DeepSeek V3.1 (Reasoning) vs Moonshot v1

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

Quick Verdict

Moonshot v1 wins overall with a score of 44 vs 25 (19 point difference).Moonshot v1 wins 4 out of 4 categories.

Knowledge

Moonshot v1

DeepSeek V3.1 (Reasoning)

31.8

Moonshot v1

50.8

34
MMLU
53
33
GPQA
52
31
SuperGPQA
50
29
OpenBookQA
48

Coding

Moonshot v1

DeepSeek V3.1 (Reasoning)

26

Moonshot v1

45

26
HumanEval
45

Mathematics

Moonshot v1

DeepSeek V3.1 (Reasoning)

33

Moonshot v1

52

34
AIME 2023
53
36
AIME 2024
55
35
AIME 2025
54
30
HMMT Feb 2023
49
32
HMMT Feb 2024
51
31
HMMT Feb 2025
50
33
BRUMO 2025
52

Reasoning

Moonshot v1

DeepSeek V3.1 (Reasoning)

31

Moonshot v1

50

32
SimpleQA
51
30
MuSR
49

Frequently Asked Questions

Which is better, DeepSeek V3.1 (Reasoning) or Moonshot v1?

Moonshot v1 scores higher overall with 44 vs 25, a difference of 19 points across all benchmarks.

Which is better for knowledge tasks, DeepSeek V3.1 (Reasoning) or Moonshot v1?

Moonshot v1 leads in knowledge tasks with an average score of 50.8 vs 31.8.

Which is better for coding, DeepSeek V3.1 (Reasoning) or Moonshot v1?

Moonshot v1 leads in coding with an average score of 45 vs 26.

Which is better for math, DeepSeek V3.1 (Reasoning) or Moonshot v1?

Moonshot v1 leads in math with an average score of 52 vs 33.

Which is better for reasoning, DeepSeek V3.1 (Reasoning) or Moonshot v1?

Moonshot v1 leads in reasoning with an average score of 50 vs 31.