GeekRadar / The comparison desk
COMPARE.
AI models
anthropic

Claude 3 Haiku (20240307)

Model evaluation

Capability & preference

Human preference · Arena

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

Claude 3 Haiku (20240307) 1194.6 Claude 3 Haiku (20240307) · 95% interval 1190.9–1198.4 · 117,701 votes · 2026-11-06

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 · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,154.79 rating

Confidence interval: 1,145.87 – 1,163.72

16707 votes

Category
chinese
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / codinghuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,198.72 rating

Confidence interval: 1,191.89 – 1,205.56

20898 votes

Category
coding
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / creative_writinghuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,157.01 rating

Confidence interval: 1,149.47 – 1,164.56

16718 votes

Category
creative_writing
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / englishhuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,206.10 rating

Confidence interval: 1,201.26 – 1,210.95

64537 votes

Category
english
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / exclude_tieshuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,075.01 rating

Confidence interval: 1,069.20 – 1,080.83

77578 votes

Category
exclude_ties
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / experthuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,147.66 rating

Confidence interval: 1,136.88 – 1,158.45

6336 votes

Category
expert
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / frenchhuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,194.80 rating

Confidence interval: 1,174.02 – 1,215.59

1772 votes

Category
french
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / germanhuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,173.62 rating

Confidence interval: 1,159.36 – 1,187.87

3508 votes

Category
german
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / hard_promptshuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,173.73 rating

Confidence interval: 1,168.10 – 1,179.36

33698 votes

Category
hard_prompts
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / hard_prompts_englishhuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,181.28 rating

Confidence interval: 1,174.61 – 1,187.95

20673 votes

Category
hard_prompts_english
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,157.42 rating

Confidence interval: 1,149.27 – 1,165.57

13237 votes

Category
industry_business_and_management_and_financial_operations
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,151.83 rating

Confidence interval: 1,144.61 – 1,159.04

17134 votes

Category
industry_entertainment_and_sports_and_media
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / industry_legal_and_governmenthuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,191.41 rating

Confidence interval: 1,180.62 – 1,202.20

6799 votes

Category
industry_legal_and_government
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,178.30 rating

Confidence interval: 1,171.08 – 1,185.51

20565 votes

Category
industry_life_and_physical_and_social_science
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / industry_mathematicalhuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,187.46 rating

Confidence interval: 1,179.65 – 1,195.27

13491 votes

Category
industry_mathematical
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,165.11 rating

Confidence interval: 1,153.96 – 1,176.27

6385 votes

Category
industry_medicine_and_healthcare
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / industry_software_and_it_serviceshuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,193.83 rating

Confidence interval: 1,187.92 – 1,199.74

34258 votes

Category
industry_software_and_it_services
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,193.82 rating

Confidence interval: 1,187.57 – 1,200.07

29876 votes

Category
industry_writing_and_literature_and_language
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / instruction_followinghuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,173.19 rating

Confidence interval: 1,167.91 – 1,178.47

43031 votes

Category
instruction_following
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / japanesehuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,102.41 rating

Confidence interval: 1,085.43 – 1,119.39

2063 votes

Category
japanese
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / koreanhuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,108.72 rating

Confidence interval: 1,089.87 – 1,127.56

2364 votes

Category
korean
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / longer_queryhuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,190.08 rating

Confidence interval: 1,182.48 – 1,197.69

13976 votes

Category
longer_query
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / mathhuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,188.06 rating

Confidence interval: 1,180.64 – 1,195.48

14983 votes

Category
math
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / multi_turnhuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,189.84 rating

Confidence interval: 1,182.62 – 1,197.06

19344 votes

Category
multi_turn
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / non_englishhuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,177.40 rating

Confidence interval: 1,172.37 – 1,182.43

53164 votes

Category
non_english
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / overallhuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,194.64 rating

Confidence interval: 1,190.91 – 1,198.38

117701 votes

Category
overall
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / polishhuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,233.42 rating

Confidence interval: 1,182.33 – 1,284.51

264 votes

Category
polish
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / russianhuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,204.10 rating

Confidence interval: 1,195.95 – 1,212.26

11956 votes

Category
russian
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference
Arena text preference / spanishhuman preference · 2026-11-06 · rating

Claude 3 Haiku (20240307)

1,165.56 rating

Confidence interval: 1,144.99 – 1,186.13

1704 votes

Category
spanish
Evaluated configuration
Claude 3 Haiku (20240307)
Evaluation type
human preference

Run it locally

Inference · memory estimate

Claude 3 Haiku (20240307)

Not enough verified architecture information for a reliable VRAM estimate.

Capabilities

Model license
Proprietary