GeekRadar / The comparison desk
COMPARE.
AI models
openai

o3 (2025-04-16)

Model evaluation

Capability & preference

Human preference · Arena

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

o3 (2025-04-16) 1409.8 o3 (2025-04-16) · 95% interval 1406.2–1413.4 · 58,901 votes · 2026-10-28

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 · 2026-10-28 · rating

o3 (2025-04-16)

1,437.01 rating

Confidence interval: 1,427.07 – 1,446.94

3892 votes

Category
chinese
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / codinghuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,407.37 rating

Confidence interval: 1,401.19 – 1,413.54

11581 votes

Category
coding
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / creative_writinghuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,358.42 rating

Confidence interval: 1,351.14 – 1,365.69

7636 votes

Category
creative_writing
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / englishhuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,416.47 rating

Confidence interval: 1,411.78 – 1,421.17

27278 votes

Category
english
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / exclude_tieshuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,393.02 rating

Confidence interval: 1,388.08 – 1,397.95

42877 votes

Category
exclude_ties
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / experthuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,402.13 rating

Confidence interval: 1,390.89 – 1,413.38

3019 votes

Category
expert
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / frenchhuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,430.51 rating

Confidence interval: 1,408.48 – 1,452.53

826 votes

Category
french
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / germanhuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,420.10 rating

Confidence interval: 1,404.08 – 1,436.12

1510 votes

Category
german
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / hard_promptshuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,402.08 rating

Confidence interval: 1,397.31 – 1,406.84

25528 votes

Category
hard_prompts
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / hard_prompts_englishhuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,410.70 rating

Confidence interval: 1,404.58 – 1,416.81

12104 votes

Category
hard_prompts_english
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,392.52 rating

Confidence interval: 1,385.71 – 1,399.33

9481 votes

Category
industry_business_and_management_and_financial_operations
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,372.76 rating

Confidence interval: 1,366.34 – 1,379.18

10511 votes

Category
industry_entertainment_and_sports_and_media
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,428.03 rating

Confidence interval: 1,417.86 – 1,438.21

3745 votes

Category
industry_legal_and_government
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,429.36 rating

Confidence interval: 1,422.65 – 1,436.07

9926 votes

Category
industry_life_and_physical_and_social_science
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / industry_mathematicalhuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,422.58 rating

Confidence interval: 1,412.31 – 1,432.85

3455 votes

Category
industry_mathematical
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,447.56 rating

Confidence interval: 1,436.53 – 1,458.59

3328 votes

Category
industry_medicine_and_healthcare
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,414.65 rating

Confidence interval: 1,409.43 – 1,419.87

19144 votes

Category
industry_software_and_it_services
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,371.31 rating

Confidence interval: 1,365.39 – 1,377.22

13240 votes

Category
industry_writing_and_literature_and_language
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / instruction_followinghuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,367.81 rating

Confidence interval: 1,362.27 – 1,373.34

15501 votes

Category
instruction_following
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / japanesehuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,403.50 rating

Confidence interval: 1,386.05 – 1,420.94

1418 votes

Category
japanese
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / koreanhuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,369.50 rating

Confidence interval: 1,351.29 – 1,387.72

1202 votes

Category
korean
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / longer_queryhuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,371.54 rating

Confidence interval: 1,365.42 – 1,377.66

11538 votes

Category
longer_query
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / mathhuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,426.31 rating

Confidence interval: 1,416.36 – 1,436.26

3699 votes

Category
math
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / multi_turnhuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,404.72 rating

Confidence interval: 1,397.94 – 1,411.49

9654 votes

Category
multi_turn
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / non_englishhuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,401.15 rating

Confidence interval: 1,396.67 – 1,405.64

31571 votes

Category
non_english
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / overallhuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,409.84 rating

Confidence interval: 1,406.23 – 1,413.45

58901 votes

Category
overall
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / polishhuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,437.86 rating

Confidence interval: 1,427.12 – 1,448.60

3834 votes

Category
polish
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / russianhuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,406.30 rating

Confidence interval: 1,397.55 – 1,415.05

4677 votes

Category
russian
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference
Arena text preference / spanishhuman preference · 2026-10-28 · rating

o3 (2025-04-16)

1,394.85 rating

Confidence interval: 1,376.94 – 1,412.75

1221 votes

Category
spanish
Evaluated configuration
o3 (2025-04-16)
Evaluation type
human preference

Run it locally

Inference · memory estimate

o3 (2025-04-16)

Not enough verified architecture information for a reliable VRAM estimate.

Capabilities

Model license
Proprietary