DeepSeek V3.1 (Reasoning) vs Qwen2.5-1M

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

Quick Verdict

Qwen2.5-1M wins overall with a score of 66 vs 25 (41 point difference).Qwen2.5-1M wins 4 out of 4 categories.

Knowledge

Qwen2.5-1M

DeepSeek V3.1 (Reasoning)

31.8

Qwen2.5-1M

81.8

34
MMLU
84
33
GPQA
83
31
SuperGPQA
81
29
OpenBookQA
79

Coding

Qwen2.5-1M

DeepSeek V3.1 (Reasoning)

26

Qwen2.5-1M

76

26
HumanEval
76

Mathematics

Qwen2.5-1M

DeepSeek V3.1 (Reasoning)

33

Qwen2.5-1M

84

34
AIME 2023
85
36
AIME 2024
87
35
AIME 2025
86
30
HMMT Feb 2023
81
32
HMMT Feb 2024
83
31
HMMT Feb 2025
82
33
BRUMO 2025
84

Reasoning

Qwen2.5-1M

DeepSeek V3.1 (Reasoning)

31

Qwen2.5-1M

80

32
SimpleQA
81
30
MuSR
79

Frequently Asked Questions

Which is better, DeepSeek V3.1 (Reasoning) or Qwen2.5-1M?

Qwen2.5-1M scores higher overall with 66 vs 25, a difference of 41 points across all benchmarks.

Which is better for knowledge tasks, DeepSeek V3.1 (Reasoning) or Qwen2.5-1M?

Qwen2.5-1M leads in knowledge tasks with an average score of 81.8 vs 31.8.

Which is better for coding, DeepSeek V3.1 (Reasoning) or Qwen2.5-1M?

Qwen2.5-1M leads in coding with an average score of 76 vs 26.

Which is better for math, DeepSeek V3.1 (Reasoning) or Qwen2.5-1M?

Qwen2.5-1M leads in math with an average score of 84 vs 33.

Which is better for reasoning, DeepSeek V3.1 (Reasoning) or Qwen2.5-1M?

Qwen2.5-1M leads in reasoning with an average score of 80 vs 31.