GeekRadar / The comparison desk

COMPARE.

AI models

microsoft

Phi-3-mini-128k-instruct

AI model

Context window: 131,072 tokens
Open weights: Yes

Capability & preference

Human preference · Arena

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

Phi-3-mini-128k-instruct 1050.4 Phi-3 Mini 128K Instruct · 95% interval 1043.1–1057.6 · 20,685 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.

Test version

Arena text preference / chinesehuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

1,016.23 rating

Confidence interval: 1,000.36 – 1,032.10

2016 votes

Oct 28, 2026

Category: chinese
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / codinghuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

1,039.15 rating

Confidence interval: 1,026.35 – 1,051.95

3886 votes

Oct 28, 2026

Category: coding
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / creative_writinghuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

1,023.89 rating

Confidence interval: 1,009.97 – 1,037.82

2949 votes

Oct 28, 2026

Category: creative_writing
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / englishhuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

1,082.93 rating

Confidence interval: 1,074.00 – 1,091.86

12596 votes

Oct 28, 2026

Category: english
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / exclude_tieshuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

844.53 rating

Confidence interval: 832.00 – 857.05

13845 votes

Oct 28, 2026

Category: exclude_ties
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / experthuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

983.72 rating

Confidence interval: 963.53 – 1,003.91

1095 votes

Oct 28, 2026

Category: expert
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / frenchhuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

1,039.29 rating

Confidence interval: 1,004.22 – 1,074.37

372 votes

Oct 28, 2026

Category: french
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / germanhuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

1,006.05 rating

Confidence interval: 981.37 – 1,030.73

727 votes

Oct 28, 2026

Category: german
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / hard_promptshuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

1,027.89 rating

Confidence interval: 1,017.17 – 1,038.60

6167 votes

Oct 28, 2026

Category: hard_prompts
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / hard_prompts_englishhuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

1,051.13 rating

Confidence interval: 1,038.74 – 1,063.51

4149 votes

Oct 28, 2026

Category: hard_prompts_english
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

990.00 rating

Confidence interval: 974.12 – 1,005.87

2085 votes

Oct 28, 2026

Category: industry_business_and_management_and_financial_operations
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

1,009.01 rating

Confidence interval: 995.16 – 1,022.85

2922 votes

Oct 28, 2026

Category: industry_entertainment_and_sports_and_media
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

1,006.22 rating

Confidence interval: 984.23 – 1,028.22

1033 votes

Oct 28, 2026

Category: industry_legal_and_government
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

1,022.71 rating

Confidence interval: 1,009.15 – 1,036.28

3426 votes

Oct 28, 2026

Category: industry_life_and_physical_and_social_science
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / industry_mathematicalhuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

1,082.74 rating

Confidence interval: 1,069.05 – 1,096.44

2569 votes

Oct 28, 2026

Category: industry_mathematical
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

995.77 rating

Confidence interval: 974.34 – 1,017.20

1114 votes

Oct 28, 2026

Category: industry_medicine_and_healthcare
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

1,033.62 rating

Confidence interval: 1,022.39 – 1,044.85

6056 votes

Oct 28, 2026

Category: industry_software_and_it_services
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

1,040.25 rating

Confidence interval: 1,028.61 – 1,051.90

4627 votes

Oct 28, 2026

Category: industry_writing_and_literature_and_language
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / instruction_followinghuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

1,022.68 rating

Confidence interval: 1,012.57 – 1,032.78

7368 votes

Oct 28, 2026

Category: instruction_following
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / japanesehuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

899.25 rating

Confidence interval: 867.32 – 931.18

385 votes

Oct 28, 2026

Category: japanese
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / koreanhuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

855.99 rating

Confidence interval: 816.14 – 895.85

260 votes

Oct 28, 2026

Category: korean
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / longer_queryhuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

995.44 rating

Confidence interval: 979.67 – 1,011.21

2074 votes

Oct 28, 2026

Category: longer_query
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / mathhuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

1,089.01 rating

Confidence interval: 1,075.87 – 1,102.15

2813 votes

Oct 28, 2026

Category: math
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / multi_turnhuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

988.89 rating

Confidence interval: 973.37 – 1,004.40

2507 votes

Oct 28, 2026

Category: multi_turn
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / non_englishhuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

1,000.28 rating

Confidence interval: 990.45 – 1,010.11

8089 votes

Oct 28, 2026

Category: non_english
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / overallhuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

1,050.36 rating

Confidence interval: 1,043.15 – 1,057.58

20685 votes

Oct 28, 2026

Category: overall
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / russianhuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

1,003.72 rating

Confidence interval: 987.93 – 1,019.50

1800 votes

Oct 28, 2026

Category: russian
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / spanishhuman preference · 2026-10-28 · rating

Phi-3-mini-128k-instruct

1,058.52 rating

Confidence interval: 1,024.20 – 1,092.83

366 votes

Oct 28, 2026

Category: spanish
Evaluated configuration: Phi-3 Mini 128K Instruct
Evaluation type: human preference

Arena text preference / chinesehuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

1,016.28 rating

Confidence interval: 1,000.41 – 1,032.16

2016 votes

Oct 15, 2026

Category: chinese
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Arena text preference / codinghuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

1,039.25 rating

Confidence interval: 1,026.45 – 1,052.05

3886 votes

Oct 15, 2026

Category: coding
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Arena text preference / creative_writinghuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

1,024.05 rating

Confidence interval: 1,010.12 – 1,037.98

2949 votes

Oct 15, 2026

Category: creative_writing
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Arena text preference / englishhuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

1,082.88 rating

Confidence interval: 1,073.94 – 1,091.81

12596 votes

Oct 15, 2026

Category: english
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Arena text preference / exclude_tieshuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

844.54 rating

Confidence interval: 832.01 – 857.06

13845 votes

Oct 15, 2026

Category: exclude_ties
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Arena text preference / experthuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

983.76 rating

Confidence interval: 963.57 – 1,003.96

1095 votes

Oct 15, 2026

Category: expert
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Arena text preference / frenchhuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

1,038.97 rating

Confidence interval: 1,003.87 – 1,074.08

372 votes

Oct 15, 2026

Category: french
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Arena text preference / germanhuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

1,006.05 rating

Confidence interval: 981.41 – 1,030.69

727 votes

Oct 15, 2026

Category: german
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Arena text preference / hard_promptshuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

1,027.88 rating

Confidence interval: 1,017.16 – 1,038.59

6167 votes

Oct 15, 2026

Category: hard_prompts
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Arena text preference / hard_prompts_englishhuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

1,051.01 rating

Confidence interval: 1,038.63 – 1,063.40

4149 votes

Oct 15, 2026

Category: hard_prompts_english
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

990.04 rating

Confidence interval: 974.16 – 1,005.92

2085 votes

Oct 15, 2026

Category: industry_business_and_management_and_financial_operations
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

1,009.07 rating

Confidence interval: 995.22 – 1,022.91

2922 votes

Oct 15, 2026

Category: industry_entertainment_and_sports_and_media
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

1,005.97 rating

Confidence interval: 983.97 – 1,027.97

1033 votes

Oct 15, 2026

Category: industry_legal_and_government
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

1,022.69 rating

Confidence interval: 1,009.12 – 1,036.26

3426 votes

Oct 15, 2026

Category: industry_life_and_physical_and_social_science
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Arena text preference / industry_mathematicalhuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

1,082.74 rating

Confidence interval: 1,069.04 – 1,096.43

2569 votes

Oct 15, 2026

Category: industry_mathematical
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

995.83 rating

Confidence interval: 974.40 – 1,017.27

1114 votes

Oct 15, 2026

Category: industry_medicine_and_healthcare
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

1,033.70 rating

Confidence interval: 1,022.47 – 1,044.93

6056 votes

Oct 15, 2026

Category: industry_software_and_it_services
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

1,040.32 rating

Confidence interval: 1,028.67 – 1,051.96

4627 votes

Oct 15, 2026

Category: industry_writing_and_literature_and_language
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Arena text preference / instruction_followinghuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

1,022.69 rating

Confidence interval: 1,012.58 – 1,032.79

7368 votes

Oct 15, 2026

Category: instruction_following
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Arena text preference / japanesehuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

899.08 rating

Confidence interval: 867.19 – 930.97

385 votes

Oct 15, 2026

Category: japanese
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Arena text preference / koreanhuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

856.04 rating

Confidence interval: 816.19 – 895.88

260 votes

Oct 15, 2026

Category: korean
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Arena text preference / longer_queryhuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

995.48 rating

Confidence interval: 979.71 – 1,011.26

2074 votes

Oct 15, 2026

Category: longer_query
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Arena text preference / mathhuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

1,088.97 rating

Confidence interval: 1,075.84 – 1,102.11

2813 votes

Oct 15, 2026

Category: math
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Arena text preference / multi_turnhuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

988.88 rating

Confidence interval: 973.36 – 1,004.40

2507 votes

Oct 15, 2026

Category: multi_turn
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Arena text preference / non_englishhuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

1,000.37 rating

Confidence interval: 990.54 – 1,010.20

8089 votes

Oct 15, 2026

Category: non_english
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Arena text preference / overallhuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

1,050.35 rating

Confidence interval: 1,043.13 – 1,057.57

20685 votes

Oct 15, 2026

Category: overall
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Arena text preference / russianhuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

1,003.87 rating

Confidence interval: 988.08 – 1,019.66

1800 votes

Oct 15, 2026

Category: russian
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Arena text preference / spanishhuman preference · 2026-10-15 · rating

Phi-3-mini-128k-instruct

1,058.44 rating

Confidence interval: 1,024.09 – 1,092.78

366 votes

Oct 15, 2026

Category: spanish
Evaluated configuration: phi-3-mini-128k-instruct
Evaluation type: human preference

Run it locally

Inference · memory estimate

Phi-3-mini-128k-instruct

7.1 GiB checkpoint tensor payload · 3.82 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.

Capabilities

Context window: 131,072 tokens
Open weights: Yes
Model license: mit
Checkpoint tensor payload: 7.12 GiB

Input modalities

text

Output modalities

text

Processing & platform

Architecture: phi3