Data as of October 25, 2026 · How the score is built
Released Jun 12, 2026256K context
Kimi K2.7 Code
Decision readingKimi K2.7 Code scores 49 out of 100 and ranks #105 of 228. This profile shows 13 source-displayable benchmark rows; its strongest eligible category is Coding at #64. API pricing is $0.95 input and $4 output per million tokens, with cached input at $0.19.
Released Jun 12, 2026 — see all recent releases
Decision snapshot
Each value carries a field reference instead of floating alone. Markers compare this model with the current ranked and priced catalog; they are not absolute quality thresholds.
Capability
49/100
field median 50#105 of 228 ranked models
Public
#105of 228
Verified #59 of 87
Price
$0.95input / $4 output
input median $0.75cached $0.19 · blended $2.48
Speed
63tok/s
field median 96 tok/sFirst token 38.01 s
Context
256Ktokens
field median 256,000Reported for this model; direct source link not stored
Strongest published evidence
Coding ranks #64. Particularly well-suited for software development and code generation tasks.
Validate before choosing
13 published rows leave some tracked benchmark slots empty. Agentic is its lowest eligible category at #73.
Source-linked · 13 displayable benchmark rows
Category score record
Scores and ranks appear only where published evidence can be displayed. The table keeps the score, weight, cohort, and evidence state together.
| Category | Score | Rank | Percentile | Weight | Benchmarks | Evidence |
|---|---|---|---|---|---|---|
| AgenticRank #73 of 132Percentile 45thWeight 22%4 benchmarksVerified | 38.5 | 4 benchmarks | Verified | |||
| CodingRank #64 of 153Percentile 59thWeight 20%7 benchmarksVerified | 41.8 | 7 benchmarks | Verified | |||
| ReasoningWeight 17%0 benchmarksNot measured | Not measured | 0 benchmarks | Not measured | |||
| MultimodalWeight 12%0 benchmarksNot measured | Not measured | 0 benchmarks | Not measured | |||
| KnowledgeWeight 12%0 benchmarksNot measured | Not measured | 0 benchmarks | Not measured | |||
| MultilingualWeight 7%0 benchmarksNot measured | Not measured | 0 benchmarks | Not measured | |||
| Inst. FollowingWeight 5%0 benchmarksNot measured | Not measured | 0 benchmarks | Not measured | |||
| MathRank Not rankedWeight 5%2 benchmarksVerified | 58.3 | 2 benchmarks | Verified |
13 of 766 tracked benchmark slots have displayable evidence · bars run 0–100
Coverage detailsHow much of this is verified
Coverage is split by category so a strong number never hides a thin evidence base. Verified means the row is tied to a published source; provisional rows remain visible but separate.
- Agentic4/4 verified
- Coding7/7 verified
- ReasoningNot measured
- MultimodalNot measured
- KnowledgeNot measured
- MultilingualNot measured
- Inst. FollowingNot measured
- Math2/2 verified
Capability shape
Each axis shows percentile within that category’s eligible cohort. The comparison outline is the median of the six nearest public-score peers; a collapsed vertex means the category is not rank-eligible.
Kimi K2.7 Code category percentile values
- Agentic45th percentile
- Coding59th percentile
- ReasoningNot eligible
- MultimodalNot eligible
- KnowledgeNot eligible
- MultilingualNot eligible
- Instruction followingNot eligible
- MathNot eligible
The dashed outline is median of 6 nearest peers.
Eligible category ranks
- Agentic#73/132
- Coding#64/153
- ReasoningNot ranked
- MultimodalNot ranked
- KnowledgeNot ranked
- MultilingualNot ranked
- Inst. FollowingNot ranked
- MathNot ranked
Benchmark ledger
Coding opens by default. The marker compares each value with the best source-verified result in the catalog; provisional leaders do not set the reference. Expand the remaining categories for every published row.
Coding7 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| CursorBench 3.2 | Score49.7% | Versus best verified row Best verified: Claude Fable 5.1 · 73.4% | Gap23.7 behind | Weight10% ref. weight | Benchmark exact |
| LiveCodeBench (Vals)LiveCodeBench, Vals AI run | Score82.1% | Versus best verified row Best verified: Claude Fable 5.1 · 90.5% | Gap8.4 behind | Weight8% ref. weight | |
| Kimi Code Bench v2 | Score62.0% | Versus best verified row Best verified: Kimi K3.1 Pro · 77.1% | Gap15.1 behind | WeightDisplay only | Provider exact |
| ProgramBenchProgramBench: Can Language Models Rebuild Programs From Scratch? | Score53.6% | Versus best verified row Best verified: Claude Opus 5 · 93.0% | Gap39.4 behind | WeightDisplay only | Provider exact |
| MLS-Bench Lite | Score35.1% | Versus best verified row Best verified: Kimi K3 · 48.3% | Gap13.2 behind | WeightDisplay only | Provider exact |
| OpenHarmony BenchOpenHarmony Bench v1.0 | Score52.1% | Versus best verified row Best verified: Qwen3.8 Max · 60.8% | Gap8.7 behind | WeightDisplay only | Benchmark exact |
| SWE-bench (Vals)SWE-bench, Vals AI run | Score78.2% | Versus best verified row Best verified: Claude Opus 5 · 97.0% | Gap18.8 behind | WeightDisplay only | Verified |
Agentic4 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| MCP Atlas | Score76% | Versus best verified row Best verified: Muse Spark Open · 89.2% | Gap13.2 behind | Weight4% ref. weight | Provider exact |
| Terminal-Bench 2.1 (Vals)Terminal-Bench 2.1, Vals AI run | Score67.0% | Versus best verified row Best verified: GPT-6 Astra · 87.3% | Gap20.3 behind | Weight4% ref. weight | |
| Kimi Claw 24/7Kimi Claw 24/7 Bench | Score46.9% | Versus best verified row Best verified: Kimi K2.7 Code · 46.9% | GapBest verified | WeightDisplay only | Provider exact |
| MCP Mark VerifiedMCPMark-Verified | Score81.1% | Versus best verified row Best verified: Kimi K2.7 Code · 81.1% | GapBest verified | WeightDisplay only | Provider exact |
Math2 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| FrontierMath v2 (Tiers 1-3)FrontierMath v2 Tiers 1-3 | Score54.035% | Versus best verified row Best verified: GPT-6 Astra · 93.684% | Gap39.6 behind | WeightWeighted 30% | Benchmark exact |
| FrontierMath v2 (Tier 4)FrontierMath v2 Tier 4 | Score12.195% | Versus best verified row Best verified: GPT-6.1 Sol · 100.000% | Gap87.8 behind | WeightWeighted 10% | Benchmark exact |
Bars run 0–100; the dark tick marks the best source-verified value
All 13 rowsWhat it costs to get this score
Published API price against the public score. The x-axis uses a log scale; the dashed path marks models that are not beaten by a cheaper, higher-scoring option. Price uses average of published input and output rates.
Kimi K2.7 Code · 49.0 score · $2.48 blended per million tokens
The chart opens on the current model. Scroll horizontally to inspect the full price axis.
Horizontal: blended price per million tokens, log scale · Vertical: public score
Lineage
The sequence follows explicit supersedes links. Each score is estimated for that model; a relative can inform a sparse estimate but never sets a floor, so a newer release can score below an earlier one. Scores and prices remain blank when the corresponding public row or first-party rate is unavailable.
Code
Spec sheet
Each documented value carries its source. Missing fields stay visible as not sourced or not published, rather than disappearing from the page.
- API model ID
- Not published
- Context window
- 256K
- Maximum output
- Not sourced yet
- Knowledge cutoff
- Not sourced yet
- Input modalities
- Not sourced yet
- Output modalities
- Not sourced yet
- Parameters
- Not sourced yet
- Availability
- Not sourced yet
- Cloud regions
- Not tracked yet
- Lifecycle
- Current
- API capabilities
- Tool calling, structured outputs, and batch support are not tracked yet
- Prompt caching
- Published at $0.19 per million cached input tokens
- Self-host
- Kimi K2.7 Code needs ~640GB VRAM (8× NVIDIA H100 (80GB)).
- Rate limits
- Not tracked yet
How to read this profile
The visual layer above carries the decisions. These notes preserve the model, ranking, coverage, and family context behind the numbers.
Kimi K2.7 Code ranks #105 of 228 on the public leaderboard with a score of 49.02/100. Its source-verified position is #59 of 87.
Kimi K2.7 Code is a open weight model with a 256K context window. It uses an explicit reasoning mode, which can improve complex problem solving while adding latency and token use.
Official exact-value snapshot from MoonshotAI's Kimi K2.7 Code Hugging Face model card. BenchLM stores the coding-agent showcase rows as display-only launch evidence because several metrics are provider-run or benchmark variants that are not part of the weighted public benchmark spine.
Its explicit predecessor is Kimi K2.6. 13 of 766 tracked benchmark slots currently have displayable evidence. Missing categories stay blank.
Its strongest eligible category is Coding at #64, while its lowest eligible position is Agentic at #73. particularly well-suited for software development and code generation tasks.
Last updated October 25, 2026. Runtime fields remain blank until a sourced snapshot exists.
Deployment options
Self-host and provider-specific paths stay separate from benchmark evidence so operating constraints are visible before a score becomes the whole decision.
Where to run Kimi K2.7 Code by API
3 hosts publish a per-token price for it. OpenRouter has the lowest output rate, $3.35 per million tokens. Prices read from each host’s own pages on October 20, 2026.
| Host | Input /1M | Output /1M | Cached /1M | Context | Listed speed |
|---|---|---|---|---|---|
| OpenRouter | $0.67 | $3.35 | $0.18 | 262K | — |
| Nebius Token Factory | $0.95 | $4.00 | — | 262K | 231.26 tok/s |
| Novita AI | $0.95 | $4.00 | $0.19 | 262K | — |
Self-host vs API cost
Estimates at 50,000 req/day · 1000 tokens/req average.
Questions
How does Kimi K2.7 Code perform overall in AI benchmarks?
Kimi K2.7 Code ranks #105 out of 228 models on the public BenchAlign leaderboard, with a score of 49.02/100. Its evidence status is Supported, and this profile shows 13 source-displayable benchmark rows. The label describes evidence depth, not a provider quality claim; inspect category rows before choosing a workload.
Is Kimi K2.7 Code good for coding and programming?
Kimi K2.7 Code ranks #64 out of 153 eligible models for coding and programming, with a public category score of 41.8/100. Higher-ranked alternatives are available for workloads where this category decides the choice. Check the underlying rows before treating the aggregate as a workload guarantee.
Is Kimi K2.7 Code good for mathematics?
Kimi K2.7 Code has source-displayable benchmark coverage for mathematics, but the public category table does not assign it a rank there. The individual rows remain available for inspection. A missing category position means the evidence threshold was not met; it does not convert the model's unmeasured work into a zero.
Is Kimi K2.7 Code good for agentic tool use and computer tasks?
Kimi K2.7 Code ranks #73 out of 132 eligible models for agentic tool use and computer tasks, with a public category score of 38.5/100. Higher-ranked alternatives are available for workloads where this category decides the choice. Check the underlying rows before treating the aggregate as a workload guarantee.
Is Kimi K2.7 Code open source?
Kimi K2.7 Code is an open-weight model from Moonshot AI. Its weights can be downloaded for local or hosted deployment, subject to the published license. Open weight does not automatically mean open source: training data and training code may remain private, and commercial restrictions can still apply.
Does Kimi K2.7 Code have full benchmark coverage on BenchLM?
No. Kimi K2.7 Code currently has 30 source-displayable rows. The profile exposes published, non-generated evidence and leaves missing categories blank until an exact evaluation is available. Coverage describes how much was measured; it is not a penalty added to an individual benchmark result.
What is the context window size of Kimi K2.7 Code?
Kimi K2.7 Code has a reported context window of 256K in the exact-model catalog record. The value stays visible, but the profile marks its source link as unavailable instead of presenting it as directly documented. Maximum output length remains separate because providers often publish a different limit.