Compare · ModelsLive · 2 picked · head to head
GPT-5 vs GPT-5 Pro
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
GPT-5 Pro wins on 7/10 benchmarks
GPT-5 Pro wins 7 of 10 shared benchmarks. Leads in reasoning · math · knowledge.
Category leads
reasoning·GPT-5 Promath·GPT-5 Proknowledge·GPT-5 Procoding·GPT-5
Hype vs Reality
Attention vs performance
GPT-5
#98 by perf·no signal
GPT-5 Pro
#161 by perf·no signal
Best value
GPT-5
15.1x better value than GPT-5 Pro
GPT-5
9.5 pts/$
$5.63/M
GPT-5 Pro
0.6 pts/$
$67.50/M
Vendor risk
Who is behind the model
OpenAI
$840.0B·Tier 1
OpenAI
$840.0B·Tier 1
Head to head
10 benchmarks · 2 models
GPT-5GPT-5 Pro
ARC-AGI
GPT-5 Pro leads by +4.5
ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization.
GPT-5
65.7
GPT-5 Pro
70.2
ARC-AGI-2
GPT-5 Pro leads by +8.5
ARC-AGI-2 · the second iteration of the Abstraction and Reasoning Corpus, testing novel pattern recognition and abstract reasoning without prior training data.
GPT-5
9.9
GPT-5 Pro
18.3
FrontierMath-Tier-4-2025-07-01-Private
GPT-5 Pro leads by +2.1
FrontierMath Tier 4 (Jul 2025) · the most challenging tier of frontier mathematics, containing problems that push the absolute limits of AI mathematical reasoning.
GPT-5
12.5
GPT-5 Pro
14.6
FrontierMath-Tier-4-v2-Private
GPT-5 leads by +2.4
GPT-5
21.9
GPT-5 Pro
19.5
FrontierMath-Tiers-1-3-v2-Private
GPT-5 Pro leads by +0.4
GPT-5
55.4
GPT-5 Pro
55.8
HLE
GPT-5 Pro leads by +6.6
HLE (Humanity's Last Exam) · a reasoning benchmark designed to be the hardest public evaluation of AI. Questions span mathematics, physics, philosophy, and logic · curated to be at or beyond the frontier of human expert capability. Tested with and without tool augmentation. Claude Opus 4.7 scores 46.9% without tools and 54.7% with tools · making it one of the few benchmarks where the top score is below 60%.
GPT-5
21.6
GPT-5 Pro
28.2
Professional Reasoning · Finance
GPT-5 leads by +0.3
GPT-5
51.3
GPT-5 Pro
51.1
Professional Reasoning · Legal
GPT-5 Pro leads by +0.9
GPT-5
49.0
GPT-5 Pro
49.9
SimpleBench
GPT-5 Pro leads by +5.9
SimpleBench · tests fundamental reasoning capabilities with straightforward problems designed to expose gaps in basic logical and spatial thinking.
GPT-5
48.0
GPT-5 Pro
53.9
WeirdML
GPT-5 leads by +0.3
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
GPT-5
60.7
GPT-5 Pro
60.4
Full benchmark table
| Benchmark | GPT-5 | GPT-5 Pro |
|---|---|---|
ARC-AGI ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization. | 65.7 | 70.2 |
ARC-AGI-2 ARC-AGI-2 · the second iteration of the Abstraction and Reasoning Corpus, testing novel pattern recognition and abstract reasoning without prior training data. | 9.9 | 18.3 |
FrontierMath-Tier-4-2025-07-01-Private FrontierMath Tier 4 (Jul 2025) · the most challenging tier of frontier mathematics, containing problems that push the absolute limits of AI mathematical reasoning. | 12.5 | 14.6 |
FrontierMath-Tier-4-v2-Private | 21.9 | 19.5 |
FrontierMath-Tiers-1-3-v2-Private | 55.4 | 55.8 |
HLE HLE (Humanity's Last Exam) · a reasoning benchmark designed to be the hardest public evaluation of AI. Questions span mathematics, physics, philosophy, and logic · curated to be at or beyond the frontier of human expert capability. Tested with and without tool augmentation. Claude Opus 4.7 scores 46.9% without tools and 54.7% with tools · making it one of the few benchmarks where the top score is below 60%. | 21.6 | 28.2 |
Professional Reasoning · Finance | 51.3 | 51.1 |
Professional Reasoning · Legal | 49.0 | 49.9 |
SimpleBench SimpleBench · tests fundamental reasoning capabilities with straightforward problems designed to expose gaps in basic logical and spatial thinking. | 48.0 | 53.9 |
WeirdML WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns. | 60.7 | 60.4 |
Pricing · per 1M tokens · projected $/mo at 10M tokens