GeekRadar / The comparison desk

COMPARE.

AI models

moonshotai

Kimi K2.6

AI model Released Apr 20, 2026

Context window: 262,144 tokens
Maximum output: 262,144 tokens
Open weights: Yes

Capability & preference

General capability · ECI

A cross-benchmark capability estimate. Index points, not a percentage; intervals show uncertainty.

Kimi K2.6 151.6^* Kimi K2.6 · 95% interval 149.8–153.4

Human preference · Arena

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

Kimi K2.6 1455.6 Kimi K2.6 · 95% interval 1451.1–1460.0 · 39,978 votes · 2026-10-28

Model performance

Radar task score: four tests, equal weight. An experimental summary of these tasks, not a universal intelligence rating.

Kimi K2.6

Tested configuration Radar task score

Unrated

Older results · no current score

7 benchmark results for this configuration

Scientific reasoning

GPQA Diamond · 25%

Expert-written science questions; strict multiple-choice scoring. Chance accuracy is 25%.

Kimi K2.6 90.8^*% Approx. 95% interval: 87.4–94.1% · 2026-05-01 · Older result

Novel problem solving

Mystery Game Puzzles · 25%

Best-next-move puzzles in an undisclosed game variant. A narrow reasoning test, not general intelligence.

Kimi K2.6 18.0^*% Approx. 95% interval: 10.4–25.6% · 2026-07-17

Advanced mathematics

FrontierMath Tiers 1-3 · 25%

Private v2 mathematics problems with Python tools. Earlier versions and Tier 4 are excluded.

Kimi K2.6 57.2^*% Approx. 95% interval: 51.4–62.9% · 2026-06-10

Factual answers

SimpleQA Verified · 25%

Correct factual answers under a guessing prompt, graded by a model. Not a hallucination or abstention score.

Kimi K2.6 34.9^*% Approx. 95% interval: 31.9–37.9% · 2026-08-10

Competition mathematics

OTIS Mock AIME 2024-2025 · Separate result

Competition problems with model-extracted answers checked against an exact key. Earlier grading procedures are excluded.

Kimi K2.6 96.1^*% Approx. 95% interval: 91.5–100.0% · 2026-05-02 · Older result

Research mathematics

FrontierMath Tier 4 v2 · Separate result

The hardest private FrontierMath problems, corrected v2 set, with Python tools. Separate from Tiers 1-3 in the core score.

Kimi K2.6 25.6^*% Approx. 95% interval: 11.8–39.5% · 2026-06-10

Visual spatial reasoning

Furniture Assembly · Separate result

Assembly-error detection from photos and manuals, with zoom and Python tools. Uses model-assisted grading; not a general vision score.

Kimi K2.6 21.7^*% Approx. 95% interval: 11.4–31.9% · 2026-09-11

Bars show reported task scores. Whiskers are approximate 95% intervals where standard errors are available. Different reasoning budgets are not equal-cost tests. Missing results are never zero.

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

Kimi K2.6

1,520.41 rating

Confidence interval: 1,507.57 – 1,533.24

2313 votes

Oct 28, 2026

Category: chinese
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,488.98 rating

Confidence interval: 1,482.16 – 1,495.80

11084 votes

Oct 28, 2026

Category: coding
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,432.78 rating

Confidence interval: 1,424.59 – 1,440.98

6297 votes

Oct 28, 2026

Category: creative_writing
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,460.38 rating

Confidence interval: 1,454.78 – 1,465.98

18154 votes

Oct 28, 2026

Category: english
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,455.79 rating

Confidence interval: 1,450.06 – 1,461.51

29595 votes

Oct 28, 2026

Category: exclude_ties
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,490.92 rating

Confidence interval: 1,481.06 – 1,500.78

3851 votes

Oct 28, 2026

Category: expert
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,470.02 rating

Confidence interval: 1,453.71 – 1,486.32

1565 votes

Oct 28, 2026

Category: french
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,452.90 rating

Confidence interval: 1,431.73 – 1,474.07

742 votes

Oct 28, 2026

Category: german
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,470.54 rating

Confidence interval: 1,465.39 – 1,475.69

26045 votes

Oct 28, 2026

Category: hard_prompts
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,471.33 rating

Confidence interval: 1,464.86 – 1,477.81

12470 votes

Oct 28, 2026

Category: hard_prompts_english
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,445.03 rating

Confidence interval: 1,437.57 – 1,452.49

7923 votes

Oct 28, 2026

Category: industry_business_and_management_and_financial_operations
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,427.48 rating

Confidence interval: 1,420.00 – 1,434.96

8282 votes

Oct 28, 2026

Category: industry_entertainment_and_sports_and_media
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,461.74 rating

Confidence interval: 1,450.71 – 1,472.78

3023 votes

Oct 28, 2026

Category: industry_legal_and_government
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,467.83 rating

Confidence interval: 1,460.02 – 1,475.65

6710 votes

Oct 28, 2026

Category: industry_life_and_physical_and_social_science
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,479.67 rating

Confidence interval: 1,466.82 – 1,492.52

2198 votes

Oct 28, 2026

Category: industry_mathematical
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,456.51 rating

Confidence interval: 1,445.12 – 1,467.91

2927 votes

Oct 28, 2026

Category: industry_medicine_and_healthcare
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,484.05 rating

Confidence interval: 1,478.14 – 1,489.96

16029 votes

Oct 28, 2026

Category: industry_software_and_it_services
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,438.58 rating

Confidence interval: 1,431.56 – 1,445.60

9570 votes

Oct 28, 2026

Category: industry_writing_and_literature_and_language
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,450.88 rating

Confidence interval: 1,444.59 – 1,457.17

13184 votes

Oct 28, 2026

Category: instruction_following
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,441.94 rating

Confidence interval: 1,414.05 – 1,469.83

499 votes

Oct 28, 2026

Category: japanese
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,427.14 rating

Confidence interval: 1,404.47 – 1,449.82

734 votes

Oct 28, 2026

Category: korean
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,468.29 rating

Confidence interval: 1,462.26 – 1,474.33

16939 votes

Oct 28, 2026

Category: longer_query
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,474.16 rating

Confidence interval: 1,461.11 – 1,487.21

2075 votes

Oct 28, 2026

Category: math
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,453.18 rating

Confidence interval: 1,445.25 – 1,461.10

6934 votes

Oct 28, 2026

Category: multi_turn
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,446.26 rating

Confidence interval: 1,440.95 – 1,451.58

21824 votes

Oct 28, 2026

Category: non_english
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,455.56 rating

Confidence interval: 1,451.15 – 1,459.96

39978 votes

Oct 28, 2026

Category: overall
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,437.64 rating

Confidence interval: 1,416.82 – 1,458.47

759 votes

Oct 28, 2026

Category: polish
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,445.99 rating

Confidence interval: 1,436.65 – 1,455.33

4244 votes

Oct 28, 2026

Category: russian
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,463.38 rating

Confidence interval: 1,446.85 – 1,479.92

1372 votes

Oct 28, 2026

Category: spanish
Evaluated configuration: Kimi K2.6
Evaluation type: human preference

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

Kimi K2.6

1,520.66 rating

Confidence interval: 1,507.85 – 1,533.47

2322 votes

Oct 15, 2026

Category: chinese
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,488.76 rating

Confidence interval: 1,481.94 – 1,495.59

11091 votes

Oct 15, 2026

Category: coding
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,433.42 rating

Confidence interval: 1,425.23 – 1,441.62

6317 votes

Oct 15, 2026

Category: creative_writing
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,460.17 rating

Confidence interval: 1,454.57 – 1,465.78

18161 votes

Oct 15, 2026

Category: english
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,455.60 rating

Confidence interval: 1,449.87 – 1,461.33

29637 votes

Oct 15, 2026

Category: exclude_ties
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,490.74 rating

Confidence interval: 1,480.87 – 1,500.61

3852 votes

Oct 15, 2026

Category: expert
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,471.69 rating

Confidence interval: 1,455.39 – 1,488.00

1576 votes

Oct 15, 2026

Category: french
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,450.79 rating

Confidence interval: 1,429.59 – 1,471.99

745 votes

Oct 15, 2026

Category: german
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,470.42 rating

Confidence interval: 1,465.27 – 1,475.58

26068 votes

Oct 15, 2026

Category: hard_prompts
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,471.08 rating

Confidence interval: 1,464.60 – 1,477.56

12479 votes

Oct 15, 2026

Category: hard_prompts_english
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,444.62 rating

Confidence interval: 1,437.16 – 1,452.08

7937 votes

Oct 15, 2026

Category: industry_business_and_management_and_financial_operations
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,427.95 rating

Confidence interval: 1,420.46 – 1,435.43

8287 votes

Oct 15, 2026

Category: industry_entertainment_and_sports_and_media
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,461.29 rating

Confidence interval: 1,450.26 – 1,472.33

3028 votes

Oct 15, 2026

Category: industry_legal_and_government
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,468.10 rating

Confidence interval: 1,460.28 – 1,475.91

6717 votes

Oct 15, 2026

Category: industry_life_and_physical_and_social_science
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,479.70 rating

Confidence interval: 1,466.85 – 1,492.56

2199 votes

Oct 15, 2026

Category: industry_mathematical
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,456.56 rating

Confidence interval: 1,445.15 – 1,467.96

2931 votes

Oct 15, 2026

Category: industry_medicine_and_healthcare
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,483.92 rating

Confidence interval: 1,478.01 – 1,489.84

16046 votes

Oct 15, 2026

Category: industry_software_and_it_services
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,438.94 rating

Confidence interval: 1,431.92 – 1,445.97

9584 votes

Oct 15, 2026

Category: industry_writing_and_literature_and_language
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,451.01 rating

Confidence interval: 1,444.71 – 1,457.30

13200 votes

Oct 15, 2026

Category: instruction_following
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,443.16 rating

Confidence interval: 1,415.23 – 1,471.10

501 votes

Oct 15, 2026

Category: japanese
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,427.23 rating

Confidence interval: 1,404.50 – 1,449.96

734 votes

Oct 15, 2026

Category: korean
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,468.29 rating

Confidence interval: 1,462.26 – 1,474.33

16968 votes

Oct 15, 2026

Category: longer_query
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,474.74 rating

Confidence interval: 1,461.68 – 1,487.80

2074 votes

Oct 15, 2026

Category: math
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,453.17 rating

Confidence interval: 1,445.24 – 1,461.10

6941 votes

Oct 15, 2026

Category: multi_turn
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,446.35 rating

Confidence interval: 1,441.04 – 1,451.67

21852 votes

Oct 15, 2026

Category: non_english
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,455.42 rating

Confidence interval: 1,451.01 – 1,459.83

40013 votes

Oct 15, 2026

Category: overall
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,437.72 rating

Confidence interval: 1,416.88 – 1,458.56

759 votes

Oct 15, 2026

Category: polish
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,446.23 rating

Confidence interval: 1,436.89 – 1,455.57

4252 votes

Oct 15, 2026

Category: russian
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

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

Kimi K2.6

1,463.68 rating

Confidence interval: 1,447.09 – 1,480.28

1367 votes

Oct 15, 2026

Category: spanish
Evaluated configuration: kimi-k2.6
Evaluation type: human preference

Run it locally

Inference · memory estimate

Kimi K2.6

Not enough verified architecture information for a reliable VRAM estimate.

API cost

USD / month

Input tokens / monthOutput tokens / month

Standard text-token estimate. Excludes taxes, caching, batch discounts, reasoning and multimodal surcharges. Provider limits and regional availability may differ.

Kimi K2.6

Provider offer

Kimi K2.6 / Pricing tiers 2

Provider / scope	Rate	Currency	Amount / unit
DeepInframoonshotai/Kimi-K2.6Standard	Cached input	USD	0.15 / 1M tokens
	Input	USD	0.75 / 1M tokens
	Output	USD	3.5 / 1M tokens
SiliconFlowmoonshotai/Kimi-K2.6Standard	Cached input	CNY	1.1 / 1M tokens
	Input	CNY	6.5 / 1M tokens
	Output	CNY	27 / 1M tokens

Capabilities

Context window: 262,144 tokens
Maximum output: 262,144 tokens
Reasoning: Yes
Tool calling: Yes
Open weights: Yes
Knowledge cutoff: 2025-01

Input modalities

text
image
video

Output modalities

text

Processing & platform

Architecture: kimi_k25

* Provisional data