Compare · ModelsLive · 3 picked · head to head
Claude Opus 4.6 vs Claude Opus 4.6 (Fast) vs GPT-5 Pro
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
Claude Opus 4.6 wins on 15/17 benchmarks
Claude Opus 4.6 wins 15 of 17 shared benchmarks. Leads in knowledge · reasoning · math.
Category leads
knowledge·Claude Opus 4.6reasoning·Claude Opus 4.6arena·Claude Opus 4.6 (Fast)math·Claude Opus 4.6safety·Claude Opus 4.6agentic·Claude Opus 4.6coding·Claude Opus 4.6
Hype vs Reality
Attention vs performance
Claude Opus 4.6
#136 by perf·#9 by attention
Claude Opus 4.6 (Fast)
#170 by perf·#9 by attention
GPT-5 Pro
#173 by perf·#4 by attention
Best value
Claude Opus 4.6
5.1x better value than GPT-5 Pro
Claude Opus 4.6
3.2 pts/$
$15.00/M
Claude Opus 4.6 (Fast)
0.5 pts/$
$90.00/M
GPT-5 Pro
0.6 pts/$
$67.50/M
Vendor risk
Who is behind the model
Anthropic
$965.0B·Tier 1
Anthropic
$965.0B·Tier 1
OpenAI
$840.0B·Tier 1
Head to head
17 benchmarks · 3 models
Claude Opus 4.6Claude Opus 4.6 (Fast)GPT-5 Pro
Professional Reasoning · Finance
Claude Opus 4.6
53.3
Claude Opus 4.6 (Fast)
53.3
GPT-5 Pro
51.1
Professional Reasoning · Legal
Claude Opus 4.6
52.3
Claude Opus 4.6 (Fast)
52.3
GPT-5 Pro
49.9
ARC-AGI
Claude Opus 4.6 leads by +23.8
ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization.
Claude Opus 4.6
94.0
GPT-5 Pro
70.2
ARC-AGI-2
Claude Opus 4.6 leads by +50.8
ARC-AGI-2 · the second iteration of the Abstraction and Reasoning Corpus, testing novel pattern recognition and abstract reasoning without prior training data.
Claude Opus 4.6
69.2
GPT-5 Pro
18.3
Chatbot Arena Elo · Coding
Claude Opus 4.6 (Fast) leads by +5.2
Claude Opus 4.6
1537.0
Claude Opus 4.6 (Fast)
1542.2
Chatbot Arena Elo · Overall
Claude Opus 4.6 (Fast) leads by +6.4
Claude Opus 4.6
1497.3
Claude Opus 4.6 (Fast)
1503.7
FrontierMath-Tier-4-2025-07-01-Private
Claude Opus 4.6 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.
Claude Opus 4.6
22.9
GPT-5 Pro
14.6
FrontierMath-Tier-4-v2-Private
Claude Opus 4.6 leads by +7.3
Claude Opus 4.6
26.8
GPT-5 Pro
19.5
FrontierMath-Tiers-1-3-v2-Private
Claude Opus 4.6 leads by +10.2
Claude Opus 4.6
66.0
GPT-5 Pro
55.8
HLE
Claude Opus 4.6 leads by +2.9
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%.
Claude Opus 4.6
31.1
GPT-5 Pro
28.2
MASK
Claude Opus 4.6
96.3
Claude Opus 4.6 (Fast)
96.3
Remote Labor Index (RLI)
Claude Opus 4.6
4.2
Claude Opus 4.6 (Fast)
4.2
SWE Atlas · Codebase QnA
Claude Opus 4.6
33.3
Claude Opus 4.6 (Fast)
33.3
SWE Atlas · Test Writing
Claude Opus 4.6
36.7
Claude Opus 4.6 (Fast)
36.7
VisualToolBench (VTB)
Claude Opus 4.6
27.5
Claude Opus 4.6 (Fast)
27.5
SimpleBench
Claude Opus 4.6 leads by +7.2
SimpleBench · tests fundamental reasoning capabilities with straightforward problems designed to expose gaps in basic logical and spatial thinking.
Claude Opus 4.6
61.1
GPT-5 Pro
53.9
WeirdML
Claude Opus 4.6 leads by +17.6
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
Claude Opus 4.6
78.0
GPT-5 Pro
60.4
Full benchmark table
| Benchmark | Claude Opus 4.6 | Claude Opus 4.6 (Fast) | GPT-5 Pro |
|---|---|---|---|
Professional Reasoning · Finance | 53.3 | 53.3 | 51.1 |
Professional Reasoning · Legal | 52.3 | 52.3 | 49.9 |
ARC-AGI ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization. | 94.0 | — | 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. | 69.2 | — | 18.3 |
Chatbot Arena Elo · Coding | 1537.0 | 1542.2 | — |
Chatbot Arena Elo · Overall | 1497.3 | 1503.7 | — |
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. | 22.9 | — | 14.6 |
FrontierMath-Tier-4-v2-Private | 26.8 | — | 19.5 |
FrontierMath-Tiers-1-3-v2-Private | 66.0 | — | 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%. | 31.1 | — | 28.2 |
MASK | 96.3 | 96.3 | — |
Remote Labor Index (RLI) | 4.2 | 4.2 | — |
SWE Atlas · Codebase QnA | 33.3 | 33.3 | — |
SWE Atlas · Test Writing | 36.7 | 36.7 | — |
VisualToolBench (VTB) | 27.5 | 27.5 | — |
SimpleBench SimpleBench · tests fundamental reasoning capabilities with straightforward problems designed to expose gaps in basic logical and spatial thinking. | 61.1 | — | 53.9 |
WeirdML WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns. | 78.0 | — | 60.4 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $5.00 | $25.00 | 1.0M tokens (~500 books) | $100.00 | |
| $30.00 | $150.00 | 1.0M tokens (~500 books) | $600.00 | |
| $15.00 | $120.00 | 400K tokens (~200 books) | $412.50 |