GeekRadar / The comparison desk
COMPARE.
AI models
xai

grok-3-preview-02-24

Model evaluation

Capability & preference

Human preference · Arena

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

grok-3-preview-02-24 1425.5 grok-3-preview-02-24 · 95% interval 1421.2–1429.8 · 32,499 votes · 2026-10-15

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 · 2026-10-15 · rating

grok-3-preview-02-24

1,448.02 rating

Confidence interval: 1,434.41 – 1,461.64

1995 votes

Category
chinese
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / codinghuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,432.04 rating

Confidence interval: 1,423.74 – 1,440.34

5359 votes

Category
coding
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / creative_writinghuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,414.11 rating

Confidence interval: 1,405.00 – 1,423.21

4711 votes

Category
creative_writing
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / englishhuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,437.68 rating

Confidence interval: 1,432.24 – 1,443.12

17101 votes

Category
english
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / exclude_tieshuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,418.57 rating

Confidence interval: 1,412.52 – 1,424.62

23031 votes

Category
exclude_ties
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / experthuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,421.07 rating

Confidence interval: 1,406.06 – 1,436.08

1559 votes

Category
expert
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / frenchhuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,461.55 rating

Confidence interval: 1,430.78 – 1,492.31

393 votes

Category
french
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / germanhuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,430.78 rating

Confidence interval: 1,409.32 – 1,452.24

786 votes

Category
german
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / hard_promptshuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,433.46 rating

Confidence interval: 1,427.11 – 1,439.82

10594 votes

Category
hard_prompts
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / hard_prompts_englishhuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,441.39 rating

Confidence interval: 1,433.25 – 1,449.53

5831 votes

Category
hard_prompts_english
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,410.08 rating

Confidence interval: 1,400.42 – 1,419.73

3942 votes

Category
industry_business_and_management_and_financial_operations
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,415.89 rating

Confidence interval: 1,407.56 – 1,424.22

5679 votes

Category
industry_entertainment_and_sports_and_media
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,444.60 rating

Confidence interval: 1,430.43 – 1,458.77

1870 votes

Category
industry_legal_and_government
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,437.40 rating

Confidence interval: 1,429.20 – 1,445.61

5732 votes

Category
industry_life_and_physical_and_social_science
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / industry_mathematicalhuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,405.99 rating

Confidence interval: 1,394.26 – 1,417.71

2437 votes

Category
industry_mathematical
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,468.66 rating

Confidence interval: 1,453.40 – 1,483.92

1617 votes

Category
industry_medicine_and_healthcare
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,437.54 rating

Confidence interval: 1,430.79 – 1,444.29

9109 votes

Category
industry_software_and_it_services
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,410.92 rating

Confidence interval: 1,403.72 – 1,418.12

7743 votes

Category
industry_writing_and_literature_and_language
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / instruction_followinghuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,408.64 rating

Confidence interval: 1,402.06 – 1,415.21

9646 votes

Category
instruction_following
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / japanesehuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,386.48 rating

Confidence interval: 1,365.59 – 1,407.37

835 votes

Category
japanese
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / koreanhuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,373.32 rating

Confidence interval: 1,346.24 – 1,400.41

524 votes

Category
korean
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / longer_queryhuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,439.03 rating

Confidence interval: 1,430.38 – 1,447.69

4872 votes

Category
longer_query
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / mathhuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,390.57 rating

Confidence interval: 1,379.50 – 1,401.63

2682 votes

Category
math
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / multi_turnhuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,424.22 rating

Confidence interval: 1,415.23 – 1,433.21

4642 votes

Category
multi_turn
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / non_englishhuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,410.47 rating

Confidence interval: 1,404.98 – 1,415.95

15389 votes

Category
non_english
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / overallhuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,425.46 rating

Confidence interval: 1,421.15 – 1,429.77

32499 votes

Category
overall
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / polishhuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,433.24 rating

Confidence interval: 1,417.84 – 1,448.64

1621 votes

Category
polish
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / russianhuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,416.02 rating

Confidence interval: 1,404.92 – 1,427.13

2760 votes

Category
russian
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference
Arena text preference / spanishhuman preference · 2026-10-15 · rating

grok-3-preview-02-24

1,417.31 rating

Confidence interval: 1,387.14 – 1,447.49

376 votes

Category
spanish
Evaluated configuration
grok-3-preview-02-24
Evaluation type
human preference

Run it locally

Inference · memory estimate

grok-3-preview-02-24

Not enough verified architecture information for a reliable VRAM estimate.

Capabilities

Model license
Proprietary