DataLearner logo

GPT-5-Pro Benchmark Details

GPT-5-Pro currently shows benchmark results led by AIME2025 (1 / 216, score 100), HLE (118 / 565, score 42), GPQA Diamond (97 / 463, score 89.40).

Benchmark Results

GPT-5-Pro

Benchmark Results

Thinking
Tool usage

General Knowledge

6 evaluations
Benchmark / mode
Score
Rank/total
LiveBench
Standard Mode
70.48
35 / 117
ARC-AGI-1
Thinking Mode
70.17
85 / 147
HLE
unknown
31.64
191 / 565
HLE
Thinking Mode
30.70
197 / 565
HLE
Thinking ModeTools
42
118 / 565
ARC-AGI-2
Thinking Mode
18
86 / 136

Other

2 evaluations
Benchmark / mode
Score
Rank/total
GPQA Diamond
Thinking Mode
88.40
112 / 463
GPQA Diamond
Thinking ModeTools
89.40
97 / 463

Math and Reasoning

7 evaluations
Benchmark / mode
Score
Rank/total
AIME2025
Thinking Mode
96.70
21 / 216
AIME2025
Thinking ModeTools
100
1 / 216
FrontierMath v2
Thinking Level · High
55.79
25 / 58
IMO-ProofBench Advanced
Thinking Mode
28.60
8 / 24
FrontierMath Tier 4 v2
Thinking Level · High
19.51
28 / 42
FrontierMath - Tier 4
Standard Mode
14.60
23 / 80
FrontierMath - Tier 4
Thinking Mode
14.60
23 / 80

Common Sense Reasoning

1 evaluations
Benchmark / mode
Score
Rank/total
SimpleBench
Thinking Mode
61.60
28 / 93
Compare with other models