BenchAtlas

Rankings / Moonshot AI

Kimi K2 6

released 2026-04-20

BenchAtlas Index

as of 2026-09-11
73.0
base configuration
rank #53
10 families · 5 categories · high
68.8
base configuration
rank #80
7 families · 4 categories · medium
65.0
base configuration
rank #113
4 families · 3 categories · medium

Benchmark evidence

69 results
agentic coding
Release 2026-06-25
tasks_counted=3 · livebench_version=2026-06-25
thinking46.9 %independentLiveBench
τ²-Bench95.9 %independentArtificial Analysis
τ²-Benchno reasoning93.9 %independentArtificial Analysis
τ²-Bench
subset=banking
23.3 %independentArtificial Analysis
coding
Release 2026-06-25
tasks_counted=2 · livebench_version=2026-06-25
thinking78.6 %independentLiveBench
SciCode51.5 %independentArtificial Analysis
SciCodeno reasoning39.5 %independentArtificial Analysis
data analysis
Release 2026-06-25
tasks_counted=3 · livebench_version=2026-06-25
thinking65.1 %independentLiveBench
external indices
AA Coding Index61.8 pointsindependentArtificial Analysis
Intelligence Index v4.131.3 pointsindependentArtificial Analysis
Intelligence Index v4.1no reasoning23.6 pointsindependentArtificial Analysis
ECI151.0 pointsindependentEpoch AI Benchmarking Hub
ECIno reasoning151.0 pointsindependentEpoch AI Benchmarking Hub
Vals Index43.5 pointsindependentVals AI
factuality
SimpleQA Verified38.7 %independentEpoch AI Benchmarking Hub
2026-05-01
SimpleQA Verified34.9 %independentEpoch AI Benchmarking Hub
2026-08-10
human preference
Chinese (style control)older version
arena=text · category=chinese · style_control=true
1526.0
1511.8–1540.3
communityLMArena Leaderboard Dataset
Coding (style control)older version
arena=text · category=coding · style_control=true
1514.3
1507.2–1521.3
communityLMArena Leaderboard Dataset
Industry Software And It Services (style control)older version
arena=text · category=industry_software_and_it_services · style_control=true
1503.1
1497.0–1509.2
communityLMArena Leaderboard Dataset
Expert (style control)older version
arena=text · category=expert · style_control=true
1501.5
1491.1–1511.8
communityLMArena Leaderboard Dataset
Hard Prompts English (style control)older version
arena=text · category=hard_prompts_english · style_control=true
1487.1
1480.4–1493.7
communityLMArena Leaderboard Dataset
Hard Prompts (style control)older version
arena=text · category=hard_prompts · style_control=true
1484.9
1479.6–1490.2
communityLMArena Leaderboard Dataset
Industry Mathematical (style control)older version
arena=text · category=industry_mathematical · style_control=true
1483.1
1469.7–1496.4
communityLMArena Leaderboard Dataset
Math (style control)older version
arena=text · category=math · style_control=true
1478.4
1464.9–1491.9
communityLMArena Leaderboard Dataset
Industry Life And Physical And Social Science (style control)older version
arena=text · category=industry_life_and_physical_and_social_science · style_control=true
1477.0
1468.8–1485.1
communityLMArena Leaderboard Dataset
Longer Query (style control)older version
arena=text · category=longer_query · style_control=true
1475.8
1469.5–1482.0
communityLMArena Leaderboard Dataset
French (style control)older version
arena=text · category=french · style_control=true
1473.9
1456.8–1490.9
communityLMArena Leaderboard Dataset
Industry Medicine And Healthcare (style control)older version
arena=text · category=industry_medicine_and_healthcare · style_control=true
1470.1
1458.2–1482.1
communityLMArena Leaderboard Dataset
English (style control)older version
arena=text · category=english · style_control=true
1467.6
1461.9–1473.3
communityLMArena Leaderboard Dataset
Exclude Ties (style control)older version
arena=text · category=exclude_ties · style_control=true
1464.4
1458.5–1470.3
communityLMArena Leaderboard Dataset
Industry Legal And Government (style control)older version
arena=text · category=industry_legal_and_government · style_control=true
1461.1
1449.7–1472.6
communityLMArena Leaderboard Dataset
Overall (style control)
arena=text · category=overall · style_control=true
1460.7
1456.2–1465.3
communityLMArena Leaderboard Dataset
Multi Turn (style control)older version
arena=text · category=multi_turn · style_control=true
1459.8
1451.7–1468.0
communityLMArena Leaderboard Dataset
German (style control)older version
arena=text · category=german · style_control=true
1458.2
1436.6–1479.9
communityLMArena Leaderboard Dataset
Spanish (style control)older version
arena=text · category=spanish · style_control=true
1457.6
1440.4–1474.7
communityLMArena Leaderboard Dataset
Industry Business And Management And Financial Operations (style control)older version
arena=text · category=industry_business_and_management_and_financial_operations · style_control=true
1455.8
1448.1–1463.5
communityLMArena Leaderboard Dataset
Instruction Following (style control)older version
arena=text · category=instruction_following · style_control=true
1453.6
1447.1–1460.1
communityLMArena Leaderboard Dataset
Russian (style control)older version
arena=text · category=russian · style_control=true
1453.0
1443.1–1462.9
communityLMArena Leaderboard Dataset
Non English (style control)older version
arena=text · category=non_english · style_control=true
1449.3
1443.8–1454.7
communityLMArena Leaderboard Dataset
Polish (style control)older version
arena=text · category=polish · style_control=true
1445.4
1424.1–1466.8
communityLMArena Leaderboard Dataset
Industry Writing And Literature And Language (style control)older version
arena=text · category=industry_writing_and_literature_and_language · style_control=true
1438.2
1431.0–1445.5
communityLMArena Leaderboard Dataset
Creative Writing (style control)older version
arena=text · category=creative_writing · style_control=true
1430.8
1422.3–1439.3
communityLMArena Leaderboard Dataset
Industry Entertainment And Sports And Media (style control)older version
arena=text · category=industry_entertainment_and_sports_and_media · style_control=true
1427.3
1419.5–1435.1
communityLMArena Leaderboard Dataset
Korean (style control)older version
arena=text · category=korean · style_control=true
1421.9
1398.8–1445.0
communityLMArena Leaderboard Dataset
knowledge science
GPQA Diamond
implementation=artificial-analysis
91.1 %independentArtificial Analysis
GPQA Diamond90.8 %independentEpoch AI Benchmarking Hub
2026-05-01
GPQA Diamond
implementation=vals-ai
89.1 %independentVals AI
GPQA Diamond
implementation=artificial-analysis
no reasoning78.8 %independentArtificial Analysis
Humanity's Last Exam
implementation=artificial-analysis
37.5 %independentArtificial Analysis
Humanity's Last Exam
implementation=artificial-analysis
no reasoning19.6 %independentArtificial Analysis
MMLU-Pro
implementation=vals-ai
87.6 %independentVals AI
MMLU-Pro
implementation=vals-ai
87.6 %independentVals AI
language
Release 2026-06-25
tasks_counted=3 · livebench_version=2026-06-25
thinking75.1 %independentLiveBench
long context instruction
AA-LCR81.0 %independentArtificial Analysis
AA-LCRno reasoning69.7 %independentArtificial Analysis
IFBench76.0 %independentArtificial Analysis
IFBenchno reasoning44.3 %independentArtificial Analysis
Release 2026-06-25
tasks_counted=4 · livebench_version=2026-06-25
thinking64.4 %independentLiveBench
multimodal
MMMU
implementation=vals-ai
86.3 %independentVals AI
MMMU
implementation=vals-ai
86.3 %independentVals AI
professional
CaseLaw61.2 %independentVals AI
CorpFin66.7 %independentVals AI
LegalBench84.7 %independentVals AI
TaxEval74.7 %independentVals AI
reasoning math
FrontierMath39.0 %independentEpoch AI Benchmarking Hub
2026-05-07
FrontierMath Tier 4older version14.6 %independentEpoch AI Benchmarking Hub
2026-05-08
Release 2026-06-25
tasks_counted=4 · livebench_version=2026-06-25
thinking84.3 %independentLiveBench
Release 2026-06-25
tasks_counted=4 · livebench_version=2026-06-25
thinking79.4 %independentLiveBench
OTIS Mock AIME 2024–202596.1 %independentEpoch AI Benchmarking Hub
2026-05-02

Agent + model results

systems, not bare-model scores
agent + model Epoch Inspect harness + Kimi K2 6SWE-bench Verified76.7 %independentEpoch AI Benchmarking Hub
agent + model Artificial Analysis harness + Kimi K2 6Terminal-Bench 2.165.9 %independentArtificial Analysis
agent + model Artificial Analysis harness + Kimi K2 6Terminal-Bench Hard43.9 %independentArtificial Analysis
agent + model Artificial Analysis harness + Kimi K2 6Terminal-Bench Hard37.9 %independentArtificial Analysis

These scores measure the whole agent system (scaffold, tools, budgets) — they are never merged into the bare model’s numbers.

Related variants

Kimi K2 5Kimi K2