DataLearner logo

DeepSeek V3.2-Exp Benchmark Details

DeepSeek V3.2-Exp currently shows benchmark results led by SimpleQA (1 / 46, score 97.10), MMLU-Pro (28 / 175, score 85), Aider-Polyglot (11 / 59, score 74.20).

Benchmark Results

DeepSeek V3.2-Exp

Benchmark Results

Thinking
Tool usage

Knowledge Exams

5 evaluations
Benchmark / mode
Score
Rank/total
MMLU-Pro
Standard Mode
84
40 / 175
MMLU-Pro
Thinking Mode
85
28 / 175
HLE
Standard Mode
8.60
203 / 235
HLE
Thinking Mode
19.80
157 / 235
HLE
Thinking ModeTools
20.30
155 / 235

Scientific Reasoning

4 evaluations
Benchmark / mode
Score
Rank/total
GPQA Diamond
Standard Mode
74
161 / 253
GPQA Diamond
Thinking Mode
79.90
132 / 253
CritPt
Standard Mode
1.40
139 / 204
CritPt
Thinking Mode
1.40
139 / 204

Honesty & Factuality

1 evaluations
Benchmark / mode
Score
Rank/total
SimpleQA
Thinking ModeTools
97.10
1 / 46

Repository Engineering

1 evaluations
Benchmark / mode
Score
Rank/total
SWE-bench Verified
Thinking ModeTools
67.80
73 / 116

Algorithmic Coding

2 evaluations
Benchmark / mode
Score
Rank/total
LiveCodeBench
Standard Mode
55
88 / 125
LiveCodeBench
Thinking Mode
74.10
45 / 125

Mathematics

2 evaluations
Benchmark / mode
Score
Rank/total
AIME2025
Standard Mode
58
92 / 110
AIME2025
Thinking Mode
89.30
42 / 110

Agentic Development

6 evaluations
Benchmark / mode
Score
Rank/total
Aider-Polyglot
Standard Mode
70.20
17 / 59
Aider-Polyglot
Thinking Mode
74.20
11 / 59
Terminal-Bench
Standard ModeTools
23
30 / 35
Terminal-Bench
Thinking ModeTools
37.70
14 / 35
Terminal Bench Hard
Standard ModeTools
25
128 / 244
Terminal Bench Hard
Thinking ModeTools
31.10
105 / 244

Service Workflows

1 evaluations
Benchmark / mode
Score
Rank/total
τ²-Bench
Thinking ModeTools
66.70
28 / 44

Fact Finding

1 evaluations
Benchmark / mode
Score
Rank/total
BrowseComp
Thinking ModeTools
40.10
53 / 58

Cross-capability Suites

2 evaluations
Benchmark / mode
Score
Rank/total
LiveBench
Standard Mode
49.85
93 / 117
LiveBench
Thinking Mode
58.90
75 / 117

Capability Indices

1 evaluations
Benchmark / mode
Score
Rank/total
ECI
unknown
145.02
77 / 167
Compare with other models