Compare · ModelsLive · 2 picked · head to head

Claude Sonnet 4.6 vs Claude Opus 4.6

Side by side · benchmarks, pricing, and signals you can act on.

Winner summary

Claude Opus 4.6 wins 15 of 15 shared benchmarks. Leads in agentic · reasoning · arena.

Category leads
agentic·Claude Opus 4.6reasoning·Claude Opus 4.6arena·Claude Opus 4.6knowledge·Claude Opus 4.6math·Claude Opus 4.6coding·Claude Opus 4.6
Hype vs Reality
Claude Sonnet 4.6
#128 by perf·#18 by attention
UNDERRATED
Claude Opus 4.6
#72 by perf·#4 by attention
DESERVED
Best value
1.4x better value than Claude Opus 4.6
Claude Sonnet 4.6
5.3 pts/$
$9.00/M
Claude Opus 4.6
3.8 pts/$
$15.00/M
Vendor risk
Anthropic logo
Anthropic
$380.0B·Tier 1
Medium risk
Anthropic logo
Anthropic
$380.0B·Tier 1
Medium risk
Head to head
Claude Sonnet 4.6Claude Opus 4.6
APEX-Agents
Claude Opus 4.6 leads by +8.0
APEX-Agents · evaluates AI agents on complex, multi-step tasks requiring planning, tool use, and autonomous decision-making in realistic environments.
Claude Sonnet 4.6
23.7
Claude Opus 4.6
31.7
ARC-AGI
Claude Opus 4.6 leads by +7.5
ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization.
Claude Sonnet 4.6
86.5
Claude Opus 4.6
94.0
ARC-AGI-2
Claude Opus 4.6 leads by +8.8
ARC-AGI-2 · the second iteration of the Abstraction and Reasoning Corpus, testing novel pattern recognition and abstract reasoning without prior training data.
Claude Sonnet 4.6
60.4
Claude Opus 4.6
69.2
Chatbot Arena Elo · Coding
Claude Opus 4.6 leads by +17.3
Claude Sonnet 4.6
1521.0
Claude Opus 4.6
1538.3
Chatbot Arena Elo · Overall
Claude Opus 4.6 leads by +27.0
Claude Sonnet 4.6
1471.7
Claude Opus 4.6
1498.8
Chess Puzzles
Claude Opus 4.6 leads by +4.0
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
Claude Sonnet 4.6
13.0
Claude Opus 4.6
17.0
FrontierMath-2025-02-28-Private
Claude Opus 4.6 leads by +8.3
FrontierMath (Feb 2025) · original research-level math problems created by mathematicians, testing capabilities at the boundary of current AI mathematical reasoning.
Claude Sonnet 4.6
32.4
Claude Opus 4.6
40.7
FrontierMath-Tier-4-2025-07-01-Private
Claude Opus 4.6 leads by +14.6
FrontierMath Tier 4 (Jul 2025) · the most challenging tier of frontier mathematics, containing problems that push the absolute limits of AI mathematical reasoning.
Claude Sonnet 4.6
8.3
Claude Opus 4.6
22.9
GPQA diamond
Claude Opus 4.6 leads by +4.2
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
Claude Sonnet 4.6
83.2
Claude Opus 4.6
87.4
OTIS Mock AIME 2024-2025
Claude Opus 4.6 leads by +8.7
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
Claude Sonnet 4.6
85.8
Claude Opus 4.6
94.4
PostTrainBench
Claude Opus 4.6 leads by +8.4
Claude Sonnet 4.6
16.4
Claude Opus 4.6
24.8
SimpleQA Verified
Claude Opus 4.6 leads by +17.5
SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information.
Claude Sonnet 4.6
29.0
Claude Opus 4.6
46.5
SWE-Bench verified
Claude Opus 4.6 leads by +3.5
SWE-bench Verified · 500 human-validated tasks from 12 real Python repositories (Django, Flask, scikit-learn, sympy, and others). Each task requires the model to produce a git patch that resolves a real GitHub issue and passes the test suite. The verified subset eliminates ambiguous tasks from the original SWE-bench. Claude Mythos Preview leads at 93.9%, crossing 90% for the first time in 2026. Opus 4.6 scores 80.8%. The benchmark remains the most-cited evaluation for code-generation capability.
Claude Sonnet 4.6
75.2
Claude Opus 4.6
78.7
Terminal Bench
Claude Opus 4.6 leads by +26.4
Terminal-Bench 2.0 · evaluates AI agents on real terminal-based coding tasks · writing scripts, debugging, running tests, and managing projects entirely through command-line interaction. Tests both code quality and terminal fluency. Claude Opus 4.7 scores 69.4%, demonstrating significant agentic terminal competence.
Claude Sonnet 4.6
53.4
Claude Opus 4.6
79.8
WeirdML
Claude Opus 4.6 leads by +11.9
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
Claude Sonnet 4.6
66.1
Claude Opus 4.6
78.0
Full benchmark table
BenchmarkClaude Sonnet 4.6Claude Opus 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.
23.731.7
ARC-AGI
ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization.
86.594.0
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.
60.469.2
Chatbot Arena Elo · Coding
1521.01538.3
Chatbot Arena Elo · Overall
1471.71498.8
Chess Puzzles
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
13.017.0
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.
32.440.7
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.
8.322.9
GPQA diamond
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
83.287.4
OTIS Mock AIME 2024-2025
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
85.894.4
PostTrainBench
16.424.8
SimpleQA Verified
SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information.
29.046.5
SWE-Bench verified
SWE-bench Verified · 500 human-validated tasks from 12 real Python repositories (Django, Flask, scikit-learn, sympy, and others). Each task requires the model to produce a git patch that resolves a real GitHub issue and passes the test suite. The verified subset eliminates ambiguous tasks from the original SWE-bench. Claude Mythos Preview leads at 93.9%, crossing 90% for the first time in 2026. Opus 4.6 scores 80.8%. The benchmark remains the most-cited evaluation for code-generation capability.
75.278.7
Terminal Bench
Terminal-Bench 2.0 · evaluates AI agents on real terminal-based coding tasks · writing scripts, debugging, running tests, and managing projects entirely through command-line interaction. Tests both code quality and terminal fluency. Claude Opus 4.7 scores 69.4%, demonstrating significant agentic terminal competence.
53.479.8
WeirdML
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
66.178.0
Pricing · per 1M tokens · projected $/mo at 10M tokens
ModelInputOutputContextProjected $/mo
Anthropic logoClaude Sonnet 4.6$3.00$15.001.0M tokens (~500 books)$60.00
Anthropic logoClaude Opus 4.6$5.00$25.001.0M tokens (~500 books)$100.00