benchlm.ai

DeepSeek V3.1 (Reasoning) vs Gemini 2.5 Flash

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

Quick Verdict

Gemini 2.5 Flash wins overall with a score of 41 vs 25 (16 point difference).Gemini 2.5 Flash wins 4 out of 4 categories.

DeepSeek V3.1 (Reasoning)

CreatorDeepSeek
TypeOpen Weight
Context128K
Score25

Gemini 2.5 Flash

CreatorGoogle
TypeProprietary
Context1M
Score41

Knowledge

Gemini 2.5 Flash

DeepSeek V3.1 (Reasoning)

31.8

Gemini 2.5 Flash

47.8

34
MMLU
50
33
GPQA
49
31
SuperGPQA
47
29
OpenBookQA
45

Coding

Gemini 2.5 Flash

DeepSeek V3.1 (Reasoning)

26

Gemini 2.5 Flash

42

26
HumanEval
42

Mathematics

Gemini 2.5 Flash

DeepSeek V3.1 (Reasoning)

33

Gemini 2.5 Flash

49

34
AIME 2023
50
36
AIME 2024
52
35
AIME 2025
51
30
HMMT Feb 2023
46
32
HMMT Feb 2024
48
31
HMMT Feb 2025
47
33
BRUMO 2025
49

Reasoning

Gemini 2.5 Flash

DeepSeek V3.1 (Reasoning)

31

Gemini 2.5 Flash

47

32
SimpleQA
48
30
MuSR
46

Frequently Asked Questions

Which is better, DeepSeek V3.1 (Reasoning) or Gemini 2.5 Flash?

Gemini 2.5 Flash scores higher overall with 41 vs 25, a difference of 16 points across all benchmarks.

Which is better for knowledge tasks, DeepSeek V3.1 (Reasoning) or Gemini 2.5 Flash?

Gemini 2.5 Flash leads in knowledge tasks with an average score of 47.8 vs 31.8.

Which is better for coding, DeepSeek V3.1 (Reasoning) or Gemini 2.5 Flash?

Gemini 2.5 Flash leads in coding with an average score of 42 vs 26.

Which is better for math, DeepSeek V3.1 (Reasoning) or Gemini 2.5 Flash?

Gemini 2.5 Flash leads in math with an average score of 49 vs 33.

Which is better for reasoning, DeepSeek V3.1 (Reasoning) or Gemini 2.5 Flash?

Gemini 2.5 Flash leads in reasoning with an average score of 47 vs 31.

Related Comparisons

DeepSeek V3.1 (Reasoning) vs GPT-5.4DeepSeek V3.1 (Reasoning) vs Gemini 3.1 ProDeepSeek V3.1 (Reasoning) vs Claude Opus 4.6DeepSeek V3.1 (Reasoning) vs GPT-5.3 CodexDeepSeek V3.1 (Reasoning) vs Grok 4.1DeepSeek V3.1 (Reasoning) vs GPT-5.2