DeepSeek V3.1 (Reasoning) vs Nemotron-4 15B

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

Quick Verdict

Nemotron-4 15B wins overall with a score of 45 vs 25 (20 point difference).Nemotron-4 15B wins 4 out of 4 categories.

Knowledge

Nemotron-4 15B

DeepSeek V3.1 (Reasoning)

31.8

Nemotron-4 15B

51.8

34
MMLU
54
33
GPQA
53
31
SuperGPQA
51
29
OpenBookQA
49

Coding

Nemotron-4 15B

DeepSeek V3.1 (Reasoning)

26

Nemotron-4 15B

46

26
HumanEval
46

Mathematics

Nemotron-4 15B

DeepSeek V3.1 (Reasoning)

33

Nemotron-4 15B

53

34
AIME 2023
54
36
AIME 2024
56
35
AIME 2025
55
30
HMMT Feb 2023
50
32
HMMT Feb 2024
52
31
HMMT Feb 2025
51
33
BRUMO 2025
53

Reasoning

Nemotron-4 15B

DeepSeek V3.1 (Reasoning)

31

Nemotron-4 15B

51

32
SimpleQA
52
30
MuSR
50

Frequently Asked Questions

Which is better, DeepSeek V3.1 (Reasoning) or Nemotron-4 15B?

Nemotron-4 15B scores higher overall with 45 vs 25, a difference of 20 points across all benchmarks.

Which is better for knowledge tasks, DeepSeek V3.1 (Reasoning) or Nemotron-4 15B?

Nemotron-4 15B leads in knowledge tasks with an average score of 51.8 vs 31.8.

Which is better for coding, DeepSeek V3.1 (Reasoning) or Nemotron-4 15B?

Nemotron-4 15B leads in coding with an average score of 46 vs 26.

Which is better for math, DeepSeek V3.1 (Reasoning) or Nemotron-4 15B?

Nemotron-4 15B leads in math with an average score of 53 vs 33.

Which is better for reasoning, DeepSeek V3.1 (Reasoning) or Nemotron-4 15B?

Nemotron-4 15B leads in reasoning with an average score of 51 vs 31.