GeekRadar / The comparison desk
COMPARE.
AI models
moonshotai

Kimi K2 Thinking Turbo

AI model Released

Context window
262,144 tokens
Maximum output
262,144 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.

Kimi K2 Thinking Turbo 1414.4 Kimi K2 Thinking Turbo · 95% interval 1411.3–1417.4 · 63,310 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

Kimi K2 Thinking Turbo

1,456.36 rating

Confidence interval: 1,447.43 – 1,465.29

4696 votes

Category
chinese
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / codinghuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,453.55 rating

Confidence interval: 1,448.16 – 1,458.94

15253 votes

Category
coding
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / creative_writinghuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,373.26 rating

Confidence interval: 1,366.72 – 1,379.80

9523 votes

Category
creative_writing
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / englishhuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,426.26 rating

Confidence interval: 1,421.94 – 1,430.58

26349 votes

Category
english
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / exclude_tieshuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,400.03 rating

Confidence interval: 1,395.81 – 1,404.26

45409 votes

Category
exclude_ties
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / experthuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,438.77 rating

Confidence interval: 1,429.92 – 1,447.62

4697 votes

Category
expert
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / frenchhuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,425.75 rating

Confidence interval: 1,409.87 – 1,441.62

1625 votes

Category
french
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / germanhuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,389.28 rating

Confidence interval: 1,371.45 – 1,407.11

1146 votes

Category
german
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / hard_promptshuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,427.76 rating

Confidence interval: 1,423.89 – 1,431.64

37045 votes

Category
hard_prompts
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / hard_prompts_englishhuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,433.04 rating

Confidence interval: 1,427.73 – 1,438.35

15624 votes

Category
hard_prompts_english
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,397.50 rating

Confidence interval: 1,391.65 – 1,403.35

12141 votes

Category
industry_business_and_management_and_financial_operations
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,377.24 rating

Confidence interval: 1,371.34 – 1,383.14

11929 votes

Category
industry_entertainment_and_sports_and_media
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,415.77 rating

Confidence interval: 1,406.97 – 1,424.58

4569 votes

Category
industry_legal_and_government
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,418.44 rating

Confidence interval: 1,412.21 – 1,424.66

10296 votes

Category
industry_life_and_physical_and_social_science
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / industry_mathematicalhuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,435.59 rating

Confidence interval: 1,424.94 – 1,446.24

3206 votes

Category
industry_mathematical
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,416.61 rating

Confidence interval: 1,406.96 – 1,426.26

4024 votes

Category
industry_medicine_and_healthcare
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,446.21 rating

Confidence interval: 1,441.63 – 1,450.79

23155 votes

Category
industry_software_and_it_services
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,380.34 rating

Confidence interval: 1,374.89 – 1,385.79

14641 votes

Category
industry_writing_and_literature_and_language
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / instruction_followinghuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,403.12 rating

Confidence interval: 1,398.19 – 1,408.06

18708 votes

Category
instruction_following
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / japanesehuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,358.39 rating

Confidence interval: 1,332.90 – 1,383.88

570 votes

Category
japanese
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / koreanhuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,329.17 rating

Confidence interval: 1,309.85 – 1,348.50

1013 votes

Category
korean
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / longer_queryhuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,415.02 rating

Confidence interval: 1,410.08 – 1,419.95

19901 votes

Category
longer_query
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / mathhuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,428.99 rating

Confidence interval: 1,419.47 – 1,438.50

3919 votes

Category
math
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / multi_turnhuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,414.07 rating

Confidence interval: 1,407.89 – 1,420.25

10977 votes

Category
multi_turn
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / non_englishhuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,397.47 rating

Confidence interval: 1,393.65 – 1,401.29

36958 votes

Category
non_english
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / overallhuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,414.36 rating

Confidence interval: 1,411.29 – 1,417.43

63310 votes

Category
overall
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / polishhuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,395.59 rating

Confidence interval: 1,381.52 – 1,409.66

1706 votes

Category
polish
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / russianhuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,390.40 rating

Confidence interval: 1,383.40 – 1,397.40

7445 votes

Category
russian
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / spanishhuman preference · 2026-10-28 · rating

Kimi K2 Thinking Turbo

1,406.32 rating

Confidence interval: 1,392.48 – 1,420.16

1995 votes

Category
spanish
Evaluated configuration
Kimi K2 Thinking Turbo
Evaluation type
human preference
Arena text preference / chinesehuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,456.26 rating

Confidence interval: 1,447.33 – 1,465.19

4703 votes

Category
chinese
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / codinghuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,453.65 rating

Confidence interval: 1,448.26 – 1,459.04

15278 votes

Category
coding
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / creative_writinghuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,373.64 rating

Confidence interval: 1,367.10 – 1,380.17

9534 votes

Category
creative_writing
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / englishhuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,426.37 rating

Confidence interval: 1,422.05 – 1,430.69

26366 votes

Category
english
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / exclude_tieshuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,400.23 rating

Confidence interval: 1,396.00 – 1,404.46

45455 votes

Category
exclude_ties
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / experthuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,438.71 rating

Confidence interval: 1,429.86 – 1,447.56

4705 votes

Category
expert
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / frenchhuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,425.37 rating

Confidence interval: 1,409.47 – 1,441.27

1633 votes

Category
french
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / germanhuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,389.78 rating

Confidence interval: 1,371.92 – 1,407.63

1148 votes

Category
german
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / hard_promptshuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,427.89 rating

Confidence interval: 1,424.01 – 1,431.77

37086 votes

Category
hard_prompts
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / hard_prompts_englishhuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,433.12 rating

Confidence interval: 1,427.81 – 1,438.43

15636 votes

Category
hard_prompts_english
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,397.53 rating

Confidence interval: 1,391.67 – 1,403.38

12151 votes

Category
industry_business_and_management_and_financial_operations
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,377.35 rating

Confidence interval: 1,371.44 – 1,383.25

11942 votes

Category
industry_entertainment_and_sports_and_media
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,416.27 rating

Confidence interval: 1,407.46 – 1,425.08

4570 votes

Category
industry_legal_and_government
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,418.50 rating

Confidence interval: 1,412.28 – 1,424.72

10301 votes

Category
industry_life_and_physical_and_social_science
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / industry_mathematicalhuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,435.48 rating

Confidence interval: 1,424.82 – 1,446.13

3207 votes

Category
industry_mathematical
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,416.78 rating

Confidence interval: 1,407.13 – 1,426.43

4027 votes

Category
industry_medicine_and_healthcare
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,446.35 rating

Confidence interval: 1,441.77 – 1,450.94

23184 votes

Category
industry_software_and_it_services
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,380.68 rating

Confidence interval: 1,375.22 – 1,386.13

14654 votes

Category
industry_writing_and_literature_and_language
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / instruction_followinghuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,403.13 rating

Confidence interval: 1,398.19 – 1,408.06

18730 votes

Category
instruction_following
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / japanesehuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,358.24 rating

Confidence interval: 1,332.70 – 1,383.77

570 votes

Category
japanese
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / koreanhuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,329.66 rating

Confidence interval: 1,310.28 – 1,349.04

1012 votes

Category
korean
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / longer_queryhuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,415.12 rating

Confidence interval: 1,410.18 – 1,420.06

19929 votes

Category
longer_query
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / mathhuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,429.02 rating

Confidence interval: 1,419.49 – 1,438.54

3918 votes

Category
math
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / multi_turnhuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,414.18 rating

Confidence interval: 1,408.00 – 1,420.37

10982 votes

Category
multi_turn
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / non_englishhuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,397.65 rating

Confidence interval: 1,393.83 – 1,401.47

36999 votes

Category
non_english
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / overallhuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,414.49 rating

Confidence interval: 1,411.42 – 1,417.56

63368 votes

Category
overall
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / polishhuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,395.48 rating

Confidence interval: 1,381.40 – 1,409.56

1704 votes

Category
polish
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / russianhuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,390.61 rating

Confidence interval: 1,383.61 – 1,397.61

7452 votes

Category
russian
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference
Arena text preference / spanishhuman preference · 2026-10-15 · rating

Kimi K2 Thinking Turbo

1,405.66 rating

Confidence interval: 1,391.78 – 1,419.53

1993 votes

Category
spanish
Evaluated configuration
kimi-k2-thinking-turbo
Evaluation type
human preference

Run it locally

Inference · memory estimate

Kimi K2 Thinking Turbo

Not enough verified architecture information for a reliable VRAM estimate.

Capabilities

Context window
262,144 tokens
Maximum output
262,144 tokens
Reasoning
Yes
Tool calling
Yes
Open weights
Yes
Knowledge cutoff
2024-08
Model license
Modified MIT

Input modalities

  • text

Output modalities

  • text