GeekRadar / The comparison desk
COMPARE.
AI models
anthropic

Claude Sonnet 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 Sonnet 4 (20250514) 1339.3 Claude Sonnet 4 (20250514) · 95% interval 1334.9–1343.7 · 39,126 votes · 2026-11-09

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 · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,340.85 rating

Confidence interval: 1,329.21 – 1,352.49

2840 votes

Category
chinese
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / codinghuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,379.20 rating

Confidence interval: 1,371.80 – 1,386.60

7299 votes

Category
coding
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / creative_writinghuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,342.03 rating

Confidence interval: 1,333.12 – 1,350.94

4902 votes

Category
creative_writing
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / englishhuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,344.11 rating

Confidence interval: 1,338.47 – 1,349.75

17186 votes

Category
english
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / exclude_tieshuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,295.11 rating

Confidence interval: 1,289.31 – 1,300.90

28566 votes

Category
exclude_ties
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / experthuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,335.35 rating

Confidence interval: 1,322.11 – 1,348.59

2086 votes

Category
expert
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / frenchhuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,362.43 rating

Confidence interval: 1,337.13 – 1,387.73

568 votes

Category
french
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / germanhuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,330.86 rating

Confidence interval: 1,312.07 – 1,349.65

1079 votes

Category
german
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / hard_promptshuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,354.81 rating

Confidence interval: 1,349.22 – 1,360.40

17408 votes

Category
hard_prompts
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / hard_prompts_englishhuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,358.61 rating

Confidence interval: 1,351.34 – 1,365.89

7892 votes

Category
hard_prompts_english
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,318.19 rating

Confidence interval: 1,310.19 – 1,326.20

6532 votes

Category
industry_business_and_management_and_financial_operations
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,320.20 rating

Confidence interval: 1,312.42 – 1,327.98

6837 votes

Category
industry_entertainment_and_sports_and_media
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / industry_legal_and_governmenthuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,341.77 rating

Confidence interval: 1,329.65 – 1,353.89

2620 votes

Category
industry_legal_and_government
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,340.19 rating

Confidence interval: 1,332.12 – 1,348.27

6383 votes

Category
industry_life_and_physical_and_social_science
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / industry_mathematicalhuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,357.94 rating

Confidence interval: 1,345.59 – 1,370.30

2316 votes

Category
industry_mathematical
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,330.31 rating

Confidence interval: 1,317.31 – 1,343.30

2364 votes

Category
industry_medicine_and_healthcare
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / industry_software_and_it_serviceshuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,360.70 rating

Confidence interval: 1,354.60 – 1,366.80

13131 votes

Category
industry_software_and_it_services
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,350.77 rating

Confidence interval: 1,343.68 – 1,357.86

8490 votes

Category
industry_writing_and_literature_and_language
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / instruction_followinghuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,351.91 rating

Confidence interval: 1,345.25 – 1,358.58

9902 votes

Category
instruction_following
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / japanesehuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,301.98 rating

Confidence interval: 1,282.02 – 1,321.95

969 votes

Category
japanese
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / koreanhuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,290.69 rating

Confidence interval: 1,270.07 – 1,311.31

923 votes

Category
korean
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / longer_queryhuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,379.37 rating

Confidence interval: 1,372.08 – 1,386.65

7694 votes

Category
longer_query
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / mathhuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,359.11 rating

Confidence interval: 1,347.38 – 1,370.83

2445 votes

Category
math
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / multi_turnhuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,365.05 rating

Confidence interval: 1,357.13 – 1,372.97

6810 votes

Category
multi_turn
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / non_englishhuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,330.05 rating

Confidence interval: 1,324.74 – 1,335.37

21931 votes

Category
non_english
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / overallhuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,339.30 rating

Confidence interval: 1,334.87 – 1,343.73

39126 votes

Category
overall
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / polishhuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,321.53 rating

Confidence interval: 1,309.56 – 1,333.50

2913 votes

Category
polish
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / russianhuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,352.54 rating

Confidence interval: 1,342.00 – 1,363.08

3212 votes

Category
russian
Evaluated configuration
Claude Sonnet 4 (20250514)
Evaluation type
human preference
Arena text preference / spanishhuman preference · 2026-11-09 · rating

Claude Sonnet 4 (20250514)

1,349.78 rating

Confidence interval: 1,329.20 – 1,370.35

889 votes

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

Run it locally

Inference · memory estimate

Claude Sonnet 4 (20250514)

Not enough verified architecture information for a reliable VRAM estimate.

Capabilities

Model license
Proprietary