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-4-fast-chat
1,457.12 rating
Confidence interval: 1,428.75 – 1,485.49
463 votes
Category
chinese
Evaluated configuration
grok-4-fast-chat
Evaluation type
human preference
Arena text preference / codinghuman preference · 2026-10-15 · rating
grok-4-fast-chat
1,428.66 rating
Confidence interval: 1,411.98 – 1,445.34
1183 votes
Category
coding
Evaluated configuration
grok-4-fast-chat
Evaluation type
human preference
Arena text preference / creative_writinghuman preference · 2026-10-15 · rating
grok-4-fast-chat
1,386.38 rating
Confidence interval: 1,366.42 – 1,406.34
889 votes
Category
creative_writing
Evaluated configuration
grok-4-fast-chat
Evaluation type
human preference
Arena text preference / englishhuman preference · 2026-10-15 · rating
grok-4-fast-chat
1,413.78 rating
Confidence interval: 1,402.44 – 1,425.13
2643 votes
Category
english
Evaluated configuration
grok-4-fast-chat
Evaluation type
human preference
Arena text preference / exclude_tieshuman preference · 2026-10-15 · rating
grok-4-fast-chat
1,390.40 rating
Confidence interval: 1,379.74 – 1,401.06
4551 votes
Category
exclude_ties
Evaluated configuration
grok-4-fast-chat
Evaluation type
human preference
Arena text preference / experthuman preference · 2026-10-15 · rating
grok-4-fast-chat
1,398.45 rating
Confidence interval: 1,364.30 – 1,432.60
288 votes
Category
expert
Evaluated configuration
grok-4-fast-chat
Evaluation type
human preference
Arena text preference / hard_promptshuman preference · 2026-10-15 · rating
grok-4-fast-chat
1,412.10 rating
Confidence interval: 1,401.42 – 1,422.78
3011 votes
Category
hard_prompts
Evaluated configuration
grok-4-fast-chat
Evaluation type
human preference
Arena text preference / hard_prompts_englishhuman preference · 2026-10-15 · rating
grok-4-fast-chat
1,417.78 rating
Confidence interval: 1,401.31 – 1,434.26
1267 votes
Category
hard_prompts_english
Evaluated configuration
grok-4-fast-chat
Evaluation type
human preference
Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-15 · rating