Compare · ModelsLive · 3 picked · head to head
GPT-5.4 Pro vs GPT-5.5 vs GPT-5.5 Pro
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
GPT-5.5 Pro wins on 7/14 benchmarks
GPT-5.5 Pro wins 7 of 14 shared benchmarks. Leads in reasoning · math.
Category leads
reasoning·GPT-5.5 Proknowledge·GPT-5.5math·GPT-5.5 Progeneral·GPT-5.5coding·GPT-5.5
Hype vs Reality
Attention vs performance
GPT-5.4 Pro
#40 by perf·#9 by attention
GPT-5.5
#64 by perf·#9 by attention
GPT-5.5 Pro
#8 by perf·#9 by attention
Best value
GPT-5.5
4.6x better value than GPT-5.5 Pro
GPT-5.4 Pro
0.6 pts/$
$105.00/M
GPT-5.5
3.3 pts/$
$17.50/M
GPT-5.5 Pro
0.7 pts/$
$105.00/M
Vendor risk
Who is behind the model
OpenAI
$840.0B·Tier 1
OpenAI
$840.0B·Tier 1
OpenAI
$840.0B·Tier 1
Head to head
14 benchmarks · 3 models
GPT-5.4 ProGPT-5.5GPT-5.5 Pro
ARC-AGI
GPT-5.5 Pro leads by +1.5
ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization.
GPT-5.4 Pro
94.5
GPT-5.5
95.0
GPT-5.5 Pro
96.5
ARC-AGI-2
GPT-5.5 leads by +0.4
ARC-AGI-2 · the second iteration of the Abstraction and Reasoning Corpus, testing novel pattern recognition and abstract reasoning without prior training data.
GPT-5.4 Pro
83.3
GPT-5.5
85.0
GPT-5.5 Pro
84.6
Chess Puzzles
GPT-5.5 leads by +1.0
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
GPT-5.4 Pro
58.6
GPT-5.5
61.1
GPT-5.5 Pro
60.0
FrontierMath-2025-02-28-Private
GPT-5.5 Pro leads by +0.7
FrontierMath (Feb 2025) · original research-level math problems created by mathematicians, testing capabilities at the boundary of current AI mathematical reasoning.
GPT-5.4 Pro
50.0
GPT-5.5
51.7
GPT-5.5 Pro
52.4
FrontierMath-Tier-4-2025-07-01-Private
GPT-5.5 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-5.4 Pro
37.5
GPT-5.5
35.4
GPT-5.5 Pro
39.6
FrontierMath-Tier-4-v2-Private
GPT-5.5 Pro leads by +5.5
GPT-5.4 Pro
58.5
GPT-5.5
72.5
GPT-5.5 Pro
78.0
FrontierMath-Tiers-1-3-v2-Private
GPT-5.5 Pro leads by +2.6
GPT-5.4 Pro
86.8
GPT-5.5
89.8
GPT-5.5 Pro
92.3
GPQA diamond
GPT-5.5 leads by +0.1
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
GPT-5.4 Pro
92.8
GPT-5.5
95.8
GPT-5.5 Pro
95.7
SimpleBench
GPT-5.5 Pro leads by +3.4
SimpleBench · tests fundamental reasoning capabilities with straightforward problems designed to expose gaps in basic logical and spatial thinking.
GPT-5.4 Pro
68.9
GPT-5.5
62.8
GPT-5.5 Pro
72.3
SimpleQA Verified
GPT-5.5 Pro leads by +1.5
SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information.
GPT-5.4 Pro
47.8
GPT-5.5
63.0
GPT-5.5 Pro
64.5
Dtbench
GPT-5.5
93.3
GPT-5.5 Pro
93.3
Lmca
GPT-5.5 leads by +0.4
GPT-5.5
63.9
GPT-5.5 Pro
63.5
OTIS Mock AIME 2024-2025
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
GPT-5.5
100.0
GPT-5.5 Pro
100.0
WeirdML
GPT-5.5 leads by +27.5
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
GPT-5.4 Pro
57.4
GPT-5.5
84.9
Full benchmark table
| Benchmark | GPT-5.4 Pro | GPT-5.5 | GPT-5.5 Pro |
|---|---|---|---|
ARC-AGI ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization. | 94.5 | 95.0 | 96.5 |
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. | 83.3 | 85.0 | 84.6 |
Chess Puzzles Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities. | 58.6 | 61.1 | 60.0 |
FrontierMath-2025-02-28-Private FrontierMath (Feb 2025) · original research-level math problems created by mathematicians, testing capabilities at the boundary of current AI mathematical reasoning. | 50.0 | 51.7 | 52.4 |
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. | 37.5 | 35.4 | 39.6 |
FrontierMath-Tier-4-v2-Private | 58.5 | 72.5 | 78.0 |
FrontierMath-Tiers-1-3-v2-Private | 86.8 | 89.8 | 92.3 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 92.8 | 95.8 | 95.7 |
SimpleBench SimpleBench · tests fundamental reasoning capabilities with straightforward problems designed to expose gaps in basic logical and spatial thinking. | 68.9 | 62.8 | 72.3 |
SimpleQA Verified SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information. | 47.8 | 63.0 | 64.5 |
Dtbench | n/a | 93.3 | 93.3 |
Lmca | n/a | 63.9 | 63.5 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | n/a | 100.0 | 100.0 |
WeirdML WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns. | 57.4 | 84.9 | n/a |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $30.00 | $180.00 | 1.1M tokens (~525 books) | $675.00 | |
| $5.00 | $30.00 | 1.1M tokens (~525 books) | $112.50 | |
| $30.00 | $180.00 | 1.1M tokens (~525 books) | $675.00 |