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
Phi-3-medium-4k-instruct
1,108.06 rating
Confidence interval: 1,095.21 – 1,120.90
2819 votes
Category
chinese
Evaluated configuration
Phi-3 Medium 4K Instruct
Evaluation type
human preference
Arena text preference / codinghuman preference · 2026-10-28 · rating
Phi-3-medium-4k-instruct
1,129.84 rating
Confidence interval: 1,119.67 – 1,140.02
3973 votes
Category
coding
Evaluated configuration
Phi-3 Medium 4K Instruct
Evaluation type
human preference
Arena text preference / creative_writinghuman preference · 2026-10-28 · rating
Phi-3-medium-4k-instruct
1,106.17 rating
Confidence interval: 1,095.41 – 1,116.93
3735 votes
Category
creative_writing
Evaluated configuration
Phi-3 Medium 4K Instruct
Evaluation type
human preference
Arena text preference / englishhuman preference · 2026-10-28 · rating
Phi-3-medium-4k-instruct
1,169.26 rating
Confidence interval: 1,162.67 – 1,175.85
13283 votes
Category
english
Evaluated configuration
Phi-3 Medium 4K Instruct
Evaluation type
human preference
Arena text preference / exclude_tieshuman preference · 2026-10-28 · rating
Phi-3-medium-4k-instruct
978.43 rating
Confidence interval: 969.92 – 986.95
15851 votes
Category
exclude_ties
Evaluated configuration
Phi-3 Medium 4K Instruct
Evaluation type
human preference
Arena text preference / experthuman preference · 2026-10-28 · rating
Phi-3-medium-4k-instruct
1,107.92 rating
Confidence interval: 1,090.14 – 1,125.69
1091 votes
Category
expert
Evaluated configuration
Phi-3 Medium 4K Instruct
Evaluation type
human preference
Arena text preference / frenchhuman preference · 2026-10-28 · rating
Phi-3-medium-4k-instruct
1,110.78 rating
Confidence interval: 1,076.45 – 1,145.12
350 votes
Category
french
Evaluated configuration
Phi-3 Medium 4K Instruct
Evaluation type
human preference
Arena text preference / germanhuman preference · 2026-10-28 · rating
Phi-3-medium-4k-instruct
1,100.76 rating
Confidence interval: 1,079.32 – 1,122.20
851 votes
Category
german
Evaluated configuration
Phi-3 Medium 4K Instruct
Evaluation type
human preference
Arena text preference / hard_promptshuman preference · 2026-10-28 · rating
Phi-3-medium-4k-instruct
1,126.44 rating
Confidence interval: 1,118.40 – 1,134.49
6823 votes
Category
hard_prompts
Evaluated configuration
Phi-3 Medium 4K Instruct
Evaluation type
human preference
Arena text preference / hard_prompts_englishhuman preference · 2026-10-28 · rating
Phi-3-medium-4k-instruct
1,142.42 rating
Confidence interval: 1,132.63 – 1,152.21
4087 votes
Category
hard_prompts_english
Evaluated configuration
Phi-3 Medium 4K Instruct
Evaluation type
human preference
Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-28 · rating
Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-15 · rating
Phi-3-medium-4k-instruct
1,081.17 rating
Confidence interval: 1,070.72 – 1,091.62
3888 votes
Category
industry_entertainment_and_sports_and_media
Evaluated configuration
phi-3-medium-4k-instruct
Evaluation type
human preference
Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-15 · rating
Phi-3-medium-4k-instruct
1,130.55 rating
Confidence interval: 1,114.22 – 1,146.89
1474 votes
Category
industry_legal_and_government
Evaluated configuration
phi-3-medium-4k-instruct
Evaluation type
human preference
Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-15 · rating
Phi-3-medium-4k-instruct
1,137.86 rating
Confidence interval: 1,127.26 – 1,148.46
4186 votes
Category
industry_life_and_physical_and_social_science
Evaluated configuration
phi-3-medium-4k-instruct
Evaluation type
human preference
Arena text preference / industry_mathematicalhuman preference · 2026-10-15 · rating
Phi-3-medium-4k-instruct
1,175.77 rating
Confidence interval: 1,164.57 – 1,186.97
2729 votes
Category
industry_mathematical
Evaluated configuration
phi-3-medium-4k-instruct
Evaluation type
human preference
Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-15 · rating
Phi-3-medium-4k-instruct
1,118.74 rating
Confidence interval: 1,101.54 – 1,135.95
1320 votes
Category
industry_medicine_and_healthcare
Evaluated configuration
phi-3-medium-4k-instruct
Evaluation type
human preference
Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-15 · rating
Phi-3-medium-4k-instruct
1,133.36 rating
Confidence interval: 1,124.90 – 1,141.81
6654 votes
Category
industry_software_and_it_services
Evaluated configuration
phi-3-medium-4k-instruct
Evaluation type
human preference
Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-15 · rating
Phi-3-medium-4k-instruct
1,126.38 rating
Confidence interval: 1,117.79 – 1,134.98
6655 votes
Category
industry_writing_and_literature_and_language
Evaluated configuration
phi-3-medium-4k-instruct
Evaluation type
human preference
Arena text preference / instruction_followinghuman preference · 2026-10-15 · rating
Phi-3-medium-4k-instruct
1,113.29 rating
Confidence interval: 1,105.97 – 1,120.62
9385 votes
Category
instruction_following
Evaluated configuration
phi-3-medium-4k-instruct
Evaluation type
human preference
Arena text preference / japanesehuman preference · 2026-10-15 · rating
Phi-3-medium-4k-instruct
1,042.36 rating
Confidence interval: 1,018.14 – 1,066.58
683 votes
Category
japanese
Evaluated configuration
phi-3-medium-4k-instruct
Evaluation type
human preference
Arena text preference / koreanhuman preference · 2026-10-15 · rating
Phi-3-medium-4k-instruct
954.03 rating
Confidence interval: 928.71 – 979.35
732 votes
Category
korean
Evaluated configuration
phi-3-medium-4k-instruct
Evaluation type
human preference
Arena text preference / longer_queryhuman preference · 2026-10-15 · rating
Phi-3-medium-4k-instruct
1,119.86 rating
Confidence interval: 1,106.93 – 1,132.79
2150 votes
Category
longer_query
Evaluated configuration
phi-3-medium-4k-instruct
Evaluation type
human preference
Arena text preference / mathhuman preference · 2026-10-15 · rating
Phi-3-medium-4k-instruct
1,173.35 rating
Confidence interval: 1,162.80 – 1,183.90
3238 votes
Category
math
Evaluated configuration
phi-3-medium-4k-instruct
Evaluation type
human preference
Arena text preference / multi_turnhuman preference · 2026-10-15 · rating
Phi-3-medium-4k-instruct
1,087.97 rating
Confidence interval: 1,077.02 – 1,098.91
3824 votes
Category
multi_turn
Evaluated configuration
phi-3-medium-4k-instruct
Evaluation type
human preference
Arena text preference / non_englishhuman preference · 2026-10-15 · rating
Phi-3-medium-4k-instruct
1,093.87 rating
Confidence interval: 1,086.99 – 1,100.75
11772 votes
Category
non_english
Evaluated configuration
phi-3-medium-4k-instruct
Evaluation type
human preference
Arena text preference / overallhuman preference · 2026-10-15 · rating
Phi-3-medium-4k-instruct
1,137.51 rating
Confidence interval: 1,132.39 – 1,142.63
25055 votes
Category
overall
Evaluated configuration
phi-3-medium-4k-instruct
Evaluation type
human preference
Arena text preference / russianhuman preference · 2026-10-15 · rating
Phi-3-medium-4k-instruct
1,144.93 rating
Confidence interval: 1,133.51 – 1,156.35
2964 votes
Category
russian
Evaluated configuration
phi-3-medium-4k-instruct
Evaluation type
human preference
Arena text preference / spanishhuman preference · 2026-10-15 · rating
Phi-3-medium-4k-instruct
1,094.97 rating
Confidence interval: 1,064.05 – 1,125.88
419 votes
Category
spanish
Evaluated configuration
phi-3-medium-4k-instruct
Evaluation type
human preference
Run it locally
Inference · memory estimate
Phi-3-medium-4k-instruct
26.0 GiB checkpoint tensor payload · 13.96 B stored tensor elements · BF16
Stored tensors can include buffers and repeated shared weights. A verified unique parameter count is unavailable. Checkpoint size is not the memory required to run the model.