Compare · ModelsLive · 2 picked · head to head
GPT-5.4 vs GPT-5.2
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
GPT-5.4 wins on 9/13 benchmarks
GPT-5.4 wins 9 of 13 shared benchmarks. Leads in agentic · reasoning · arena.
Category leads
agentic·GPT-5.4reasoning·GPT-5.4arena·GPT-5.4knowledge·GPT-5.2math·GPT-5.4coding·GPT-5.4
Hype vs Reality
Attention vs performance
GPT-5.4
#44 by perf·no signal
GPT-5.2
#74 by perf·no signal
Best value
GPT-5.2
1.0x better value than GPT-5.4
GPT-5.4
6.7 pts/$
$8.75/M
GPT-5.2
6.9 pts/$
$7.88/M
Vendor risk
Who is behind the model
OpenAI
$840.0B·Tier 1
OpenAI
$840.0B·Tier 1
Head to head
13 benchmarks · 2 models
GPT-5.4GPT-5.2
APEX-Agents
GPT-5.4 leads by +1.6
APEX-Agents · evaluates AI agents on complex, multi-step tasks requiring planning, tool use, and autonomous decision-making in realistic environments.
GPT-5.4
35.9
GPT-5.2
34.3
ARC-AGI
GPT-5.4 leads by +7.5
ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization.
GPT-5.4
93.7
GPT-5.2
86.2
ARC-AGI-2
GPT-5.4 leads by +21.0
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
74.0
GPT-5.2
52.9
Chatbot Arena Elo · Overall
GPT-5.4 leads by +26.3
GPT-5.4
1465.8
GPT-5.2
1439.5
Chess Puzzles
GPT-5.2 leads by +5.0
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
GPT-5.4
44.0
GPT-5.2
49.0
FrontierMath-2025-02-28-Private
GPT-5.4 leads by +6.9
FrontierMath (Feb 2025) · original research-level math problems created by mathematicians, testing capabilities at the boundary of current AI mathematical reasoning.
GPT-5.4
47.6
GPT-5.2
40.7
FrontierMath-Tier-4-2025-07-01-Private
GPT-5.4 leads by +8.3
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
27.1
GPT-5.2
18.8
GPQA diamond
GPT-5.4 leads by +2.5
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
91.1
GPT-5.2
88.5
OTIS Mock AIME 2024-2025
GPT-5.2 leads by +0.8
OTIS Mock AIME 2024–2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
GPT-5.4
95.3
GPT-5.2
96.1
PostTrainBench
GPT-5.2 leads by +1.1
GPT-5.4
20.2
GPT-5.2
21.4
SimpleQA Verified
GPT-5.4 leads by +5.9
SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information.
GPT-5.4
44.8
GPT-5.2
38.9
SWE-Bench verified
GPT-5.4 leads by +3.1
GPT-5.4
76.9
GPT-5.2
73.8
WeirdML
GPT-5.2 leads by +14.8
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
GPT-5.4
57.4
GPT-5.2
72.2
Full benchmark table
| Benchmark | GPT-5.4 | GPT-5.2 |
|---|---|---|
APEX-Agents APEX-Agents · evaluates AI agents on complex, multi-step tasks requiring planning, tool use, and autonomous decision-making in realistic environments. | 35.9 | 34.3 |
ARC-AGI ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization. | 93.7 | 86.2 |
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. | 74.0 | 52.9 |
Chatbot Arena Elo · Overall | 1465.8 | 1439.5 |
Chess Puzzles Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities. | 44.0 | 49.0 |
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. | 47.6 | 40.7 |
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. | 27.1 | 18.8 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 91.1 | 88.5 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024–2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 95.3 | 96.1 |
PostTrainBench | 20.2 | 21.4 |
SimpleQA Verified SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information. | 44.8 | 38.9 |
SWE-Bench verified | 76.9 | 73.8 |
WeirdML WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns. | 57.4 | 72.2 |
Pricing · per 1M tokens · projected $/mo at 10M tokens