Compare · ModelsLive · 2 picked · head to head
GPT-5.4 Pro vs GPT-5.6 Sol
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
GPT-5.6 Sol wins on 7/9 benchmarks
GPT-5.6 Sol wins 7 of 9 shared benchmarks. Leads in reasoning · knowledge · math.
Category leads
reasoning·GPT-5.6 Solknowledge·GPT-5.6 Solmath·GPT-5.6 Solcoding·GPT-5.6 Sol
Hype vs Reality
Attention vs performance
GPT-5.4 Pro
#37 by perf·#4 by attention
GPT-5.6 Sol
#30 by perf·#4 by attention
Best value
GPT-5.6 Sol
18.2x better value than GPT-5.4 Pro
GPT-5.4 Pro
0.6 pts/$
$105.00/M
GPT-5.6 Sol
11.1 pts/$
$6.00/M
Vendor risk
Who is behind the model
OpenAI
$840.0B·Tier 1
OpenAI
$840.0B·Tier 1
Head to head
9 benchmarks · 2 models
GPT-5.4 ProGPT-5.6 Sol
ARC-AGI
GPT-5.6 Sol leads by +3.0
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-5.6 Sol
97.5
ARC-AGI-2
GPT-5.6 Sol leads by +9.2
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-5.6 Sol
92.5
Chess Puzzles
GPT-5.6 Sol leads by +5.7
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
GPT-5.4 Pro
56.4
GPT-5.6 Sol
62.1
FrontierMath-Tier-4-v2-Private
GPT-5.6 Sol leads by +24.4
GPT-5.4 Pro
58.5
GPT-5.6 Sol
82.9
FrontierMath-Tiers-1-3-v2-Private
GPT-5.6 Sol leads by +6.7
GPT-5.4 Pro
82.5
GPT-5.6 Sol
89.1
GPQA diamond
GPT-5.4 Pro leads by +1.5
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 Pro
92.8
GPT-5.6 Sol
91.3
SimpleBench
GPT-5.4 Pro leads by +2.9
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-5.6 Sol
66.0
SimpleQA Verified
GPT-5.6 Sol leads by +23.4
SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information.
GPT-5.4 Pro
46.3
GPT-5.6 Sol
69.7
WeirdML
GPT-5.6 Sol leads by +32.0
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-5.6 Sol
89.4
Full benchmark table
| Benchmark | GPT-5.4 Pro | GPT-5.6 Sol |
|---|---|---|
ARC-AGI ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization. | 94.5 | 97.5 |
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 | 92.5 |
Chess Puzzles Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities. | 56.4 | 62.1 |
FrontierMath-Tier-4-v2-Private | 58.5 | 82.9 |
FrontierMath-Tiers-1-3-v2-Private | 82.5 | 89.1 |
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 | 91.3 |
SimpleBench SimpleBench · tests fundamental reasoning capabilities with straightforward problems designed to expose gaps in basic logical and spatial thinking. | 68.9 | 66.0 |
SimpleQA Verified SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information. | 46.3 | 69.7 |
WeirdML WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns. | 57.4 | 89.4 |
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 | |
| $2.00 | $10.00 | 1.1M tokens (~525 books) | $40.00 |