Compare · ModelsLive · 2 picked · head to head
Claude Opus 4.1 vs Grok 4
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
Grok 4 wins on 7/14 benchmarks
Grok 4 wins 7 of 14 shared benchmarks. Leads in knowledge · math · reasoning.
Category leads
knowledge·Grok 4coding·Claude Opus 4.1math·Grok 4general·Claude Opus 4.1reasoning·Grok 4
Hype vs Reality
Attention vs performance
Claude Opus 4.1
#213 by perf·#9 by attention
Grok 4
#87 by perf·#15 by attention
Vendor risk
Who is behind the model
Anthropic
$965.0B·Tier 1
xAI
$250.0B·Tier 1
Head to head
14 benchmarks · 2 models
Claude Opus 4.1Grok 4
Chess Puzzles
Grok 4 leads by +22.1
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
Claude Opus 4.1
2.1
Grok 4
24.2
Cybench
Grok 4 leads by +1.0
Cybench · evaluates AI on real Capture-The-Flag cybersecurity challenges, testing vulnerability analysis, exploitation, and security reasoning.
Claude Opus 4.1
42.0
Grok 4
43.0
DeepResearch Bench
Claude Opus 4.1 leads by +1.0
DeepResearch Bench · evaluates AI on complex multi-step research tasks requiring information gathering, synthesis, and producing comprehensive analyses.
Claude Opus 4.1
48.3
Grok 4
47.3
FrontierMath-2025-02-28-Private
Grok 4 leads by +27.2
FrontierMath (Feb 2025) · original research-level math problems created by mathematicians, testing capabilities at the boundary of current AI mathematical reasoning.
Claude Opus 4.1
7.2
Grok 4
34.5
FrontierMath-Tier-4-2025-07-01-Private
Claude Opus 4.1 leads by +0.7
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 4.1
4.2
Grok 4
3.5
Gdpval
Claude Opus 4.1 leads by +22.5
Claude Opus 4.1
43.6
Grok 4
21.1
GPQA diamond
Grok 4 leads by +13.0
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
Claude Opus 4.1
69.7
Grok 4
82.7
Lech Mazur Writing
Claude Opus 4.1 leads by +7.8
Lech Mazur Writing · evaluates creative writing ability, assessing prose quality, narrative coherence, and stylistic sophistication.
Claude Opus 4.1
84.7
Grok 4
76.9
Metr Time Horizons
Claude Opus 4.1 leads by +0.2
Claude Opus 4.1
66.8
Grok 4
66.6
OTIS Mock AIME 2024-2025
Grok 4 leads by +15.1
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
Claude Opus 4.1
68.9
Grok 4
84.0
SimpleBench
Grok 4 leads by +0.6
SimpleBench · tests fundamental reasoning capabilities with straightforward problems designed to expose gaps in basic logical and spatial thinking.
Claude Opus 4.1
52.0
Grok 4
52.6
SimpleQA Verified
Grok 4 leads by +13.1
SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information.
Claude Opus 4.1
34.8
Grok 4
47.9
Terminal Bench
Claude Opus 4.1 leads by +10.8
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 Opus 4.1
38.0
Grok 4
27.2
WeirdML
Claude Opus 4.1 leads by +0.1
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
Claude Opus 4.1
45.9
Grok 4
45.7
Full benchmark table
| Benchmark | Claude Opus 4.1 | Grok 4 |
|---|---|---|
Chess Puzzles Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities. | 2.1 | 24.2 |
Cybench Cybench · evaluates AI on real Capture-The-Flag cybersecurity challenges, testing vulnerability analysis, exploitation, and security reasoning. | 42.0 | 43.0 |
DeepResearch Bench DeepResearch Bench · evaluates AI on complex multi-step research tasks requiring information gathering, synthesis, and producing comprehensive analyses. | 48.3 | 47.3 |
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. | 7.2 | 34.5 |
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. | 4.2 | 3.5 |
Gdpval | 43.6 | 21.1 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 69.7 | 82.7 |
Lech Mazur Writing Lech Mazur Writing · evaluates creative writing ability, assessing prose quality, narrative coherence, and stylistic sophistication. | 84.7 | 76.9 |
Metr Time Horizons | 66.8 | 66.6 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 68.9 | 84.0 |
SimpleBench SimpleBench · tests fundamental reasoning capabilities with straightforward problems designed to expose gaps in basic logical and spatial thinking. | 52.0 | 52.6 |
SimpleQA Verified SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information. | 34.8 | 47.9 |
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. | 38.0 | 27.2 |
WeirdML WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns. | 45.9 | 45.7 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $15.00 | $75.00 | 200K tokens (~100 books) | $300.00 | |
| — | — | — | — |
People also compared