GeekRadar / The comparison desk
COMPARE.
AI models
anthropic

Claude Opus 4 (20250514)

Model evaluation

Capability & preference

Human preference · Arena

Overall text preferences from head-to-head votes. Not factual accuracy or a task-success rate.

Claude Opus 4 (20250514) 1366.0 Claude Opus 4 (20250514) · 95% interval 1361.7–1370.4 · 42,965 votes · 2026-11-08

Recorded results

Only the same benchmark, version, workload, scenario and unit share a row. Missing results are not zero. System results include the full submitted hardware and software configuration.

Arena text preference / chinesehuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,382.11 rating

Confidence interval: 1,371.06 – 1,393.17

3093 votes

Category
chinese
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / codinghuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,401.67 rating

Confidence interval: 1,394.43 – 1,408.91

7819 votes

Category
coding
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / creative_writinghuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,374.63 rating

Confidence interval: 1,366.04 – 1,383.22

5398 votes

Category
creative_writing
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / englishhuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,375.68 rating

Confidence interval: 1,370.16 – 1,381.20

19006 votes

Category
english
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / exclude_tieshuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,331.76 rating

Confidence interval: 1,326.07 – 1,337.45

31100 votes

Category
exclude_ties
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / experthuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,369.12 rating

Confidence interval: 1,356.66 – 1,381.57

2279 votes

Category
expert
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / frenchhuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,358.85 rating

Confidence interval: 1,334.40 – 1,383.31

630 votes

Category
french
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / germanhuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,371.95 rating

Confidence interval: 1,354.81 – 1,389.08

1274 votes

Category
german
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / hard_promptshuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,376.43 rating

Confidence interval: 1,370.98 – 1,381.87

18897 votes

Category
hard_prompts
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / hard_prompts_englishhuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,381.74 rating

Confidence interval: 1,374.66 – 1,388.82

8490 votes

Category
hard_prompts_english
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,337.10 rating

Confidence interval: 1,329.33 – 1,344.87

7033 votes

Category
industry_business_and_management_and_financial_operations
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,350.22 rating

Confidence interval: 1,342.82 – 1,357.62

7512 votes

Category
industry_entertainment_and_sports_and_media
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / industry_legal_and_governmenthuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,372.82 rating

Confidence interval: 1,361.51 – 1,384.14

2847 votes

Category
industry_legal_and_government
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,362.66 rating

Confidence interval: 1,354.85 – 1,370.47

6970 votes

Category
industry_life_and_physical_and_social_science
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / industry_mathematicalhuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,376.80 rating

Confidence interval: 1,365.29 – 1,388.31

2612 votes

Category
industry_mathematical
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,365.27 rating

Confidence interval: 1,353.10 – 1,377.44

2603 votes

Category
industry_medicine_and_healthcare
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / industry_software_and_it_serviceshuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,381.41 rating

Confidence interval: 1,375.43 – 1,387.39

14197 votes

Category
industry_software_and_it_services
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,381.41 rating

Confidence interval: 1,374.46 – 1,388.35

9284 votes

Category
industry_writing_and_literature_and_language
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / instruction_followinghuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,373.40 rating

Confidence interval: 1,366.87 – 1,379.92

10665 votes

Category
instruction_following
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / japanesehuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,329.82 rating

Confidence interval: 1,310.97 – 1,348.66

1115 votes

Category
japanese
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / koreanhuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,304.11 rating

Confidence interval: 1,284.61 – 1,323.61

1033 votes

Category
korean
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / longer_queryhuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,401.35 rating

Confidence interval: 1,394.23 – 1,408.48

8370 votes

Category
longer_query
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / mathhuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,372.98 rating

Confidence interval: 1,361.82 – 1,384.15

2735 votes

Category
math
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / multi_turnhuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,388.03 rating

Confidence interval: 1,380.36 – 1,395.70

7364 votes

Category
multi_turn
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / non_englishhuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,352.01 rating

Confidence interval: 1,346.85 – 1,357.17

23948 votes

Category
non_english
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / overallhuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,366.03 rating

Confidence interval: 1,361.69 – 1,370.36

42965 votes

Category
overall
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / polishhuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,343.65 rating

Confidence interval: 1,332.30 – 1,355.01

3158 votes

Category
polish
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / russianhuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,381.18 rating

Confidence interval: 1,370.96 – 1,391.41

3439 votes

Category
russian
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference
Arena text preference / spanishhuman preference · 2026-11-08 · rating

Claude Opus 4 (20250514)

1,356.69 rating

Confidence interval: 1,337.09 – 1,376.28

997 votes

Category
spanish
Evaluated configuration
Claude Opus 4 (20250514)
Evaluation type
human preference

Run it locally

Inference · memory estimate

Claude Opus 4 (20250514)

Not enough verified architecture information for a reliable VRAM estimate.

Capabilities

Model license
Proprietary