GeekRadar / The comparison desk
COMPARE.
AI models
xai

Grok 4 Fast (Reasoning)

Model evaluation

Capability & preference

Human preference · Arena

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

Grok 4 Fast (Reasoning) 1399.6 Grok 4 Fast (Reasoning) · 95% interval 1394.6–1404.6 · 18,270 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

Grok 4 Fast (Reasoning)

1,436.23 rating

Confidence interval: 1,419.03 – 1,453.43

1156 votes

Category
chinese
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / codinghuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,419.95 rating

Confidence interval: 1,410.62 – 1,429.28

3940 votes

Category
coding
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / creative_writinghuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,378.03 rating

Confidence interval: 1,366.25 – 1,389.82

2490 votes

Category
creative_writing
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / englishhuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,406.54 rating

Confidence interval: 1,399.64 – 1,413.44

8100 votes

Category
english
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / exclude_tieshuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,379.98 rating

Confidence interval: 1,372.96 – 1,387.00

12872 votes

Category
exclude_ties
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / experthuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,411.82 rating

Confidence interval: 1,392.85 – 1,430.80

942 votes

Category
expert
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / frenchhuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,431.98 rating

Confidence interval: 1,397.52 – 1,466.45

290 votes

Category
french
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / germanhuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,382.83 rating

Confidence interval: 1,352.57 – 1,413.10

364 votes

Category
german
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / hard_promptshuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,404.75 rating

Confidence interval: 1,398.35 – 1,411.15

9852 votes

Category
hard_prompts
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / hard_prompts_englishhuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,413.56 rating

Confidence interval: 1,404.61 – 1,422.51

4404 votes

Category
hard_prompts_english
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,394.38 rating

Confidence interval: 1,384.50 – 1,404.26

3569 votes

Category
industry_business_and_management_and_financial_operations
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,366.86 rating

Confidence interval: 1,356.73 – 1,376.98

3307 votes

Category
industry_entertainment_and_sports_and_media
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,427.48 rating

Confidence interval: 1,410.42 – 1,444.54

1145 votes

Category
industry_legal_and_government
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,426.59 rating

Confidence interval: 1,415.89 – 1,437.29

2975 votes

Category
industry_life_and_physical_and_social_science
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / industry_mathematicalhuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,401.18 rating

Confidence interval: 1,382.81 – 1,419.55

976 votes

Category
industry_mathematical
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,415.04 rating

Confidence interval: 1,395.31 – 1,434.77

873 votes

Category
industry_medicine_and_healthcare
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,416.56 rating

Confidence interval: 1,409.05 – 1,424.07

6521 votes

Category
industry_software_and_it_services
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,380.76 rating

Confidence interval: 1,371.57 – 1,389.96

4158 votes

Category
industry_writing_and_literature_and_language
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / instruction_followinghuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,382.52 rating

Confidence interval: 1,374.29 – 1,390.76

5330 votes

Category
instruction_following
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / japanesehuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,352.55 rating

Confidence interval: 1,311.74 – 1,393.35

212 votes

Category
japanese
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / koreanhuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,358.04 rating

Confidence interval: 1,327.16 – 1,388.91

370 votes

Category
korean
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / longer_queryhuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,412.18 rating

Confidence interval: 1,403.35 – 1,421.02

4529 votes

Category
longer_query
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / mathhuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,403.64 rating

Confidence interval: 1,385.90 – 1,421.39

1078 votes

Category
math
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / multi_turnhuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,404.16 rating

Confidence interval: 1,393.25 – 1,415.07

2953 votes

Category
multi_turn
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / non_englishhuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,388.48 rating

Confidence interval: 1,382.19 – 1,394.76

10133 votes

Category
non_english
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / overallhuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,399.60 rating

Confidence interval: 1,394.56 – 1,404.65

18270 votes

Category
overall
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / polishhuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,378.98 rating

Confidence interval: 1,357.38 – 1,400.57

744 votes

Category
polish
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / russianhuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,388.65 rating

Confidence interval: 1,373.61 – 1,403.69

1411 votes

Category
russian
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference
Arena text preference / spanishhuman preference · 2026-10-28 · rating

Grok 4 Fast (Reasoning)

1,404.07 rating

Confidence interval: 1,380.00 – 1,428.14

598 votes

Category
spanish
Evaluated configuration
Grok 4 Fast (Reasoning)
Evaluation type
human preference

Run it locally

Inference · memory estimate

Grok 4 Fast (Reasoning)

Not enough verified architecture information for a reliable VRAM estimate.

Capabilities

Model license
Proprietary