DeepSeek V3.1 (Reasoning) vs o1-preview

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

Quick Verdict

o1-preview wins overall with a score of 71 vs 25 (46 point difference).o1-preview wins 4 out of 4 categories.

Knowledge

o1-preview

DeepSeek V3.1 (Reasoning)

31.8

o1-preview

89

34
MMLU
92
33
GPQA
90
31
SuperGPQA
88
29
OpenBookQA
86

Coding

o1-preview

DeepSeek V3.1 (Reasoning)

26

o1-preview

86

26
HumanEval
86

Mathematics

o1-preview

DeepSeek V3.1 (Reasoning)

33

o1-preview

93

34
AIME 2023
94
36
AIME 2024
96
35
AIME 2025
95
30
HMMT Feb 2023
90
32
HMMT Feb 2024
92
31
HMMT Feb 2025
91
33
BRUMO 2025
93

Reasoning

o1-preview

DeepSeek V3.1 (Reasoning)

31

o1-preview

87

32
SimpleQA
88
30
MuSR
86

Frequently Asked Questions

Which is better, DeepSeek V3.1 (Reasoning) or o1-preview?

o1-preview scores higher overall with 71 vs 25, a difference of 46 points across all benchmarks.

Which is better for knowledge tasks, DeepSeek V3.1 (Reasoning) or o1-preview?

o1-preview leads in knowledge tasks with an average score of 89 vs 31.8.

Which is better for coding, DeepSeek V3.1 (Reasoning) or o1-preview?

o1-preview leads in coding with an average score of 86 vs 26.

Which is better for math, DeepSeek V3.1 (Reasoning) or o1-preview?

o1-preview leads in math with an average score of 93 vs 33.

Which is better for reasoning, DeepSeek V3.1 (Reasoning) or o1-preview?

o1-preview leads in reasoning with an average score of 87 vs 31.