Inference
EN
Moonshot AI

Kimi K2.7 Code

Kimi K2.7 Code is a reasoning model from Moonshot AI. 30 benchmarks count toward its index, in 6 categories.

  • Open weights
  • Reasoning
  • Input: text · image
  • Released 2026-06
RankedEnough results for an index.
IndexAll categories, out of 100.57.7±4.5Median 51.0
CoverageIndex weight that has results.85%
SpeedOutput tokens per second.35/sMedian 67/s
First tokenSeconds to the first output token.0.9 sMedian 1.3 s
Input / 1MUS dollars per 1M input tokens.$0.60Median $0.60
Output / 1MUS dollars per 1M output tokens.$3.50Median $2.55
ContextMaximum tokens in one request.262KMedian 262K
EloLMArena rating and rank.—Median 1422

Details

From OpenRouter
API IDThe model ID on OpenRouter.
moonshotai/kimi-k2.7-code
Max outputMaximum output tokens per request.
236K
Knowledge cutoffLast date of training data.
—
ToolsTool calls on OpenRouter.
Yes
JSON outputOutput follows a given JSON schema.
Yes
Cache read / 1MUS dollars per 1M cached tokens.
$0.16
WeightsLink to the published weights.
Hugging Face (opens in a new tab) ↗
RetiresDate OpenRouter removes the model.
—

Where to run

APIs that serve this model.
9 providers
ProviderTypeMaker, cloud, host or router.InputOutputContext
CoreWeaveHost$0.60$3.00262K
Deep InfraDeprecatedHost$0.68$3.40262K
OpenRouterRouter$0.60$3.50262K
SiliconFlowHost$0.859$3.80262K
Cloudflare Workers AIDeprecatedCloud$0.95$4.00262K
DatabricksDeprecatedCloud$0.95$4.00262K
Microsoft AzureCloud$0.95$4.00262K
Moonshot AIMaker$0.95$4.00262K
Novita AIHost$0.95$4.00262K

US dollars per 1M tokens, list price. From models.dev (opens in a new tab) (MIT).

Run locally

1026.88B parameters
16-bit
2157.4 GB
8-bit
1146.6 GB
4-bit
654.7 GB
Weights
Hugging Face (opens in a new tab) ↗

Memory need with a short context. The 4-bit copy needs more than 128 GB. Compare local models

Capabilities

Index per category, out of 100.
Out of 100
AgenticMulti-step tasks with tools.6 counted58.479 / 218
CodingCode writing and repair.10 counted60.468 / 209
ReasoningLogic problems and puzzles.4 counted54.686 / 227
MultimodalTasks with images and text.None counted——
KnowledgeFacts and expert knowledge.4 counted58.872 / 268
MultilingualTasks in many languages.None counted——
InstructionTasks with strict prompt rules.2 counted44.240 / 63
MathMath problems.4 counted57.058 / 168

Results

30 counted
BenchmarkThe benchmark name.ResultThe score from the source.PlacePlace among results on this benchmark.
OTIS Mock AIME 2024-2025Math95.6% ±3.140 / 200
Artificial Analysis GPQA DiamondKnowledge89.6%41 / 199
Artificial Analysis Long Context ReasoningReasoning79.3%52 / 217
τ²-Bench Tool-Agent-User EvaluationAgentic90.1%40 / 149
GPQA diamond · Epoch AIKnowledge87.9% ±2.359 / 217
Artificial Analysis Humanity's Last ExamKnowledge35.0%63 / 222
Critical Physics TasksReasoning10.0%65 / 214
SWE-benchCoding78.2% ±1.830 / 87
LiveCodeBenchCoding82.0% ±1.157 / 142
MCP AtlasAgentic76.0%21 / 52
Artificial Analysis IFBenchInstruction63.1%60 / 146
Terminal-Bench 2.1 · Vals AIAgentic67.0% ±0.436 / 75
GDPval-AA normalizedAgentic26.9%67 / 133
ProgramBench: Can Language Models Rebuild Programs From Scratch?Coding53.6%10 / 19
Artificial Analysis SciCodeCoding47.8%61 / 114
FrontierMath-Tiers-1-3-v2-PrivateMath54.0% ±3.052 / 90
LiveBench LanguageKnowledge77.9%39 / 66
Vibe Code Bench v1.1Coding47.2% ±5.273 / 123
LiveBench ReasoningReasoning82.8%49 / 80
Code MigrationCoding25.4% ±4.248 / 77
SimpleQA VerifiedKnowledge36.5% ±1.557 / 91
SkillsBenchCoding50.0% ±4.223 / 35
Terminal-Bench HardAgentic44.7%17 / 23
LiveBench CodingCoding74.0%51 / 66
LiveBench Agentic CodingAgentic45.7%52 / 66
FrontierMath-Tier-4-v2-PrivateMath12.2% ±5.258 / 72
Chess PuzzlesReasoning14.0% ±3.574 / 90
LiveBench Data AnalysisReasoning62.7%71 / 80
LiveBench MathematicsMath79.6%59 / 66
OpenHarmony Bench v1.0Coding52.1%12 / 13
LiveBench Instruction FollowingInstruction56.3%61 / 66
cursorBench32Coding49.7%17 / 18
ProgramBenchCoding0.0%34 / 65
MCPMark-VerifiedAgentic81.1%—
Kimi Code Bench v2Coding62.0%—
Kimi Claw 24/7 BenchAgentic46.9%—
MLS-Bench LiteCoding35.1%—

Sources

BenchLMCC BY-NC 4.0OpenRouterTerms restrict reuseEpoch AICC BYLiveBenchNo licence statedVals AINo licence statedmodels.devMIT

History

5 releases
Index58.4 → 57.7
23 Oct 202612 Nov 2026
Output / 1M$3.35 → $3.50
23 Oct 202612 Nov 2026

Questions

How much does Kimi K2.7 Code cost?
$0.60 per 1M input tokens, $3.50 per 1M output tokens.
How good is Kimi K2.7 Code?
Its index is 57.7 out of 100. That is place 86 of 283 ranked models.
What is Kimi K2.7 Code best at?
Knowledge: 58.8 out of 100, place 72 of 268.
How fast is Kimi K2.7 Code?
35 output tokens per second, 0.9 s to the first token.
What is the context window of Kimi K2.7 Code?
262K tokens.
Is Kimi K2.7 Code open source?
Yes. Its weights are published.
When was Kimi K2.7 Code released?
June 2026.

Badge

Kimi K2.7 Code on Inference360: place 86 of 283, index 57.7
HTML
<a href="https://inference360.ai/models/kimi-k2-7-code"><img src="https://inference360.ai/badges/kimi-k2-7-code.svg" alt="Kimi K2.7 Code on Inference360: place 86 of 283, index 57.7" height="20"></a>
Markdown
[![Kimi K2.7 Code on Inference360: place 86 of 283, index 57.7](https://inference360.ai/badges/kimi-k2-7-code.svg)](https://inference360.ai/models/kimi-k2-7-code)

Compare with

Qwen3.7 Plus57.7DeepSeek V4 Pro 042357.6Gemini 3 Pro57.9Claude Opus 4.658.0Kimi K2.658.2

Lower price

dots3-note Preview63.4 · FreeApodex 1.1 Mini55.7 · FreeGLM-5.3-Flash63.1 · $0.078/1M

More from Moonshot AI

All models
Kimi K366.9Kimi LatestUnrankedKimi K2.658.2Kimi K3.1 Pro75.7Kimi K2.5 (Reasoning)53.0