Compare · ModelsLive · 2 picked · head to head
GPT-5.4 vs Grok 4.6
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
GPT-5.4 wins on 9/17 benchmarks
GPT-5.4 wins 9 of 17 shared benchmarks. Leads in reasoning · math.
Category leads
agentic·Grok 4.6reasoning·GPT-5.4knowledge·Grok 4.6coding·Grok 4.6general·Grok 4.6math·GPT-5.4
Hype vs Reality
Attention vs performance
GPT-5.4
#104 by perf·#4 by attention
Grok 4.6
#62 by perf·#15 by attention
Best value
Grok 4.6
2.4x better value than GPT-5.4
GPT-5.4
6.0 pts/$
$8.75/M
Grok 4.6
14.4 pts/$
$4.00/M
Vendor risk
Who is behind the model
OpenAI
$840.0B·Tier 1
xAI
$250.0B·Tier 1
Head to head
17 benchmarks · 2 models
GPT-5.4Grok 4.6
APEX-Agents
Grok 4.6 leads by +12.9
APEX-Agents · evaluates AI agents on complex, multi-step tasks requiring planning, tool use, and autonomous decision-making in realistic environments.
GPT-5.4
52.4
Grok 4.6
65.3
ARC-AGI
GPT-5.4 leads by +6.2
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
Grok 4.6
87.5
ARC-AGI-2
GPT-5.4 leads by +6.9
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
Grok 4.6
67.1
Chess Puzzles
GPT-5.4 leads by +4.2
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
GPT-5.4
41.1
Grok 4.6
36.9
Deepswe
Grok 4.6 leads by +15.7
GPT-5.4
51.8
Grok 4.6
67.5
Dtbench
Grok 4.6 leads by +4.9
GPT-5.4
90.7
Grok 4.6
95.5
Ebr Bench
Grok 4.6 leads by +5.1
GPT-5.4
25.4
Grok 4.6
30.5
FrontierMath-Tier-4-v2-Private
GPT-5.4 leads by +17.3
GPT-5.4
49.0
Grok 4.6
31.7
FrontierMath-Tiers-1-3-v2-Private
GPT-5.4 leads by +12.6
GPT-5.4
78.6
Grok 4.6
66.0
Furniture Assembly
Grok 4.6 leads by +3.6
GPT-5.4
10.7
Grok 4.6
14.3
GPQA diamond
Grok 4.6 leads by +0.9
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
Grok 4.6
92.0
Lmca
GPT-5.4 leads by +4.1
GPT-5.4
61.1
Grok 4.6
57.0
Mystery Game Puzzles
GPT-5.4 leads by +3.3
GPT-5.4
30.6
Grok 4.6
27.3
OTIS Mock AIME 2024-2025
Grok 4.6 leads by +1.4
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
GPT-5.4
97.8
Grok 4.6
99.2
Proofbench
GPT-5.4 leads by +5.0
GPT-5.4
56.0
Grok 4.6
51.0
SimpleQA Verified
Grok 4.6 leads by +4.2
SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information.
GPT-5.4
45.1
Grok 4.6
49.3
WeirdML
GPT-5.4 leads by +10.4
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
GPT-5.4
77.7
Grok 4.6
67.3
Full benchmark table
| Benchmark | GPT-5.4 | Grok 4.6 |
|---|---|---|
APEX-Agents APEX-Agents · evaluates AI agents on complex, multi-step tasks requiring planning, tool use, and autonomous decision-making in realistic environments. | 52.4 | 65.3 |
ARC-AGI ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization. | 93.7 | 87.5 |
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. | 74.0 | 67.1 |
Chess Puzzles Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities. | 41.1 | 36.9 |
Deepswe | 51.8 | 67.5 |
Dtbench | 90.7 | 95.5 |
Ebr Bench | 25.4 | 30.5 |
FrontierMath-Tier-4-v2-Private | 49.0 | 31.7 |
FrontierMath-Tiers-1-3-v2-Private | 78.6 | 66.0 |
Furniture Assembly | 10.7 | 14.3 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 91.1 | 92.0 |
Lmca | 61.1 | 57.0 |
Mystery Game Puzzles | 30.6 | 27.3 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 97.8 | 99.2 |
Proofbench | 56.0 | 51.0 |
SimpleQA Verified SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information. | 45.1 | 49.3 |
WeirdML WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns. | 77.7 | 67.3 |
Pricing · per 1M tokens · projected $/mo at 10M tokens