Model record

Kimi K3

max

Moonshot AI · released Jul 16, 2026

Overall standing

Overall
61.7rank 3 of 62
61.7 is at approximately the 96th percentile of the ranked Overall field. The field spans 19.0 to 62.2, with a median of 48.9.
7 of 8 measured · 1 estimated
Category standingsReasoning 73.2Coding 71.9Math 23.4Agentic 33.4
Reasoning
73.2rank 2 of 62
73.2 is at approximately the 98th percentile of the ranked Reasoning field. The field spans 31.6 to 73.3, with a median of 64.8.
3 of 4 measured · 1 estimated
Coding
71.9rank 4 of 46
71.9 is at approximately the 92th percentile of the ranked Coding field. The field spans 10.4 to 72.4, with a median of 56.1.
2 of 2 measured
Math
23.4rank 6 of 62
23.4 is at approximately the 90th percentile of the ranked Math field. The field spans 0.0 to 32.3, with a median of 5.3.
1 of 1 measured
Agentic
33.4rank 1 of 46
33.4 is at approximately the 99th percentile of the ranked Agentic field. The field spans 3.3 to 33.4, with a median of 19.6.
1 of 1 measured
Model factsClosed weights · $3.0 / $15.0 per 1M

Copied from the linked provider page; no separate retrieval date is stored.

Provider
Moonshot AI
Released
Weights
Closed weights
Context window
1,048,576 tokens
Input price / Mtok
$3.0
Output price / Mtok
$15.0
Reasoning effort
max thinking (launch default)

Scores and sources

Measured scores retain their source and retrieval date. Missing scores are not treated as zero.

Benchmark scores, evaluation settings, sources, and retrieval dates for Kimi K3
BenchmarkCategoryScoreEvaluation settingsSource
Terminal-Bench v2.1Coding85.0Artificial Analysis independent run; 89 tasks; Terminus 2 on E2B; 3 repeats; pass@1; 250-episode cap; 2-hour timeoutOpen source (opens in a new tab)
τ³-BankingAgentic33.4Artificial Analysis independent run; 97 tasks; 5 repeats; backend-state pass@1; BM25 plus grep retrieval; 200-step capOpen source (opens in a new tab)
AA-LCRReasoning74.7Artificial Analysis independent run; 100 roughly 100k-token multi-document questions; 3 repeats; equality-checker pass@1; no toolsOpen source (opens in a new tab)
Humanity's Last ExamReasoning44.4Artificial Analysis independent run; 2,158 text-only questions; pass@1; GPT-4o (August 2024) equality checker; no toolsOpen source (opens in a new tab)
GPQA DiamondReasoning93.5Artificial Analysis independent run; 198 questions; 5 repeats; regex-graded pass@1; no toolsOpen source (opens in a new tab)
SciCodeCoding58.7Artificial Analysis independent run; 288 test subproblems; scientist background included; 3 repeats; subproblem pass@1Open source (opens in a new tab)
IFBenchReasoning—Not measured in this datasetNo source
CritPtMath23.4Artificial Analysis independent run; 70 test challenges; 5 repeats; two-step answer parsing; official CritPt grader; pass@1; no toolsOpen source (opens in a new tab)