DeepSeek V3.1 (Reasoning) vs Nemotron Ultra 253B

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

Quick Verdict

Nemotron Ultra 253B wins overall with a score of 40 vs 25 (15 point difference).Nemotron Ultra 253B wins 4 out of 4 categories.

Knowledge

Nemotron Ultra 253B

DeepSeek V3.1 (Reasoning)

31.8

Nemotron Ultra 253B

46.8

34
MMLU
49
33
GPQA
48
31
SuperGPQA
46
29
OpenBookQA
44

Coding

Nemotron Ultra 253B

DeepSeek V3.1 (Reasoning)

26

Nemotron Ultra 253B

41

26
HumanEval
41

Mathematics

Nemotron Ultra 253B

DeepSeek V3.1 (Reasoning)

33

Nemotron Ultra 253B

48

34
AIME 2023
49
36
AIME 2024
51
35
AIME 2025
50
30
HMMT Feb 2023
45
32
HMMT Feb 2024
47
31
HMMT Feb 2025
46
33
BRUMO 2025
48

Reasoning

Nemotron Ultra 253B

DeepSeek V3.1 (Reasoning)

31

Nemotron Ultra 253B

46

32
SimpleQA
47
30
MuSR
45

Frequently Asked Questions

Which is better, DeepSeek V3.1 (Reasoning) or Nemotron Ultra 253B?

Nemotron Ultra 253B scores higher overall with 40 vs 25, a difference of 15 points across all benchmarks.

Which is better for knowledge tasks, DeepSeek V3.1 (Reasoning) or Nemotron Ultra 253B?

Nemotron Ultra 253B leads in knowledge tasks with an average score of 46.8 vs 31.8.

Which is better for coding, DeepSeek V3.1 (Reasoning) or Nemotron Ultra 253B?

Nemotron Ultra 253B leads in coding with an average score of 41 vs 26.

Which is better for math, DeepSeek V3.1 (Reasoning) or Nemotron Ultra 253B?

Nemotron Ultra 253B leads in math with an average score of 48 vs 33.

Which is better for reasoning, DeepSeek V3.1 (Reasoning) or Nemotron Ultra 253B?

Nemotron Ultra 253B leads in reasoning with an average score of 46 vs 31.