LLMBenchmark.ai

Enter opens the highlighted result. Esc closes.

Kimi K2.5 vs Kimi K2.7 Code

Updated October 21, 2026. 36 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.

Replace Kimi K2.5 Replace Kimi K2.7 Code
Kimi K2.5Moonshot AI
Coding
58.5
Overall
23.5
Output $/M
$2.50
5
Kimi K2.7 CodeMoonshot AI
Coding
39.1
Overall
25.8
Output $/M
$3.35
31
vs boards won of 36 shared

Decision reading

Pick Kimi K2.5 for coding: it scores 58.5 on the Vibe Coding Index against 39.1 for Kimi K2.7 Code.

Pick Kimi K2.7 Code when what matters is LMArena WebDev or AA Coding Index, where it leads.

Model AKimi K2.5Moonshot AI58.5/1002 independent benchmarks$0.24 per solved taskModel BKimi K2.7 CodeMoonshot AI39.1/1001 independent benchmark$0.35 per solved task
Shared results
36
Kimi K2.5 only
62
Kimi K2.7 Code only
15
Like-for-like categories
7 of 9

Shared results can support a head-to-head reading. Results only one model has describe coverage, not superiority.

Which one for your work

Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.

At a glance

Coding index
Kimi K2.5 58.5 from 2 benchmarks, Kimi K2.7 Code 39.1 from 1 benchmark.
Widest category
Coding, 11.3 points apart on average over 13 shared benchmarks.
Widest single gap
Vibe Code Bench (Vals AI): Kimi K2.5 17.5% against Kimi K2.7 Code 47.2%, 29.7 points apart on a 0 to 100 scale.
Output price
Kimi K2.5 $2.50 against Kimi K2.7 Code $3.35 per million output tokens, 1.3x apart.
Context window
Both take 262K tokens.

Benchmark by benchmark

First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.

Boards won, where both have a result
CategoryKimi K2.5BarsKimi K2.7 CodeTies
Coding112
Intelligence02
Agentic work34
Reasoning03
Knowledge01
Web design08
General11
All shared boards531
CodingLike-for-likeFour of these boards feed the Vibe Coding Index112
Coding benchmarks for Kimi K2.5 and Kimi K2.7 Code
BenchmarkKimi K2.5Kimi K2.7 CodeLeader
SWE-bench Verified70.8%Not measuredOnly Kimi K2.5 measured
SWE-bench Multilingual67.3%Not measuredOnly Kimi K2.5 measured
LMArena WebDev14371474Kimi K2.7 Code
AA Coding Index46.860.8Kimi K2.7 Code
LiveCodeBench (Vals AI)83.9%82.0%Kimi K2.5
SWE-bench Verified (Vals AI)70.0%78.2%Kimi K2.7 Code
Vibe Code Bench (Vals AI)17.5%47.2%Kimi K2.7 Code
Code migration (Vals AI)7.0%25.4%Kimi K2.7 Code
ProgramBench (Vals AI)Not measured0.0%Only Kimi K2.7 Code measured
Terminal-Bench 2.1 (Vals AI)42.0%67.0%Kimi K2.7 Code
SkillsBench (Vals AI)Not measured50.0%Only Kimi K2.7 Code measured
SciCodeNot measured47.8%Only Kimi K2.7 Code measured
Terminal-Bench Hard (Artificial Analysis run)34.8%44.7%Kimi K2.7 Code
Terminal-Bench43.2%Not measuredOnly Kimi K2.5 measured
FrontierCodeNot measured30.1%Only Kimi K2.7 Code measured
WeirdML45.6%54.1%Kimi K2.7 Code
ALE-Bench821.6886.2Kimi K2.7 Code
LMArena Text: coding1483Not measuredOnly Kimi K2.5 measured
LMArena Text: software and IT1471Not measuredOnly Kimi K2.5 measured
LMArena WebDev: React14291475Kimi K2.7 Code
LMArena WebDev: HTML14261451Kimi K2.7 Code
LMArena WebDev: from an image14391513Kimi K2.7 Code
Neither model has a result on the other 16 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index02
Intelligence benchmarks for Kimi K2.5 and Kimi K2.7 Code
BenchmarkKimi K2.5Kimi K2.7 CodeLeader
AA Intelligence Index23.525.8Kimi K2.7 Code
Epoch Capabilities Index148.1150.1Kimi K2.7 Code
Agentic workLike-for-likePublic boards, shown as published, kept out of the index34
Agentic work benchmarks for Kimi K2.5 and Kimi K2.7 Code
BenchmarkKimi K2.5Kimi K2.7 CodeLeader
AA Agentic IndexNot measured21.0Only Kimi K2.7 Code measured
Design Arena: agents building web apps11361176Kimi K2.7 Code
Design Arena: agents building full-stack apps11141175Kimi K2.7 Code
Design Arena: agents building mobile apps11471143Kimi K2.5
Design Arena: agents building Android apps11241143Kimi K2.7 Code
Design Arena: agents building Godot games12191147Kimi K2.5
Design Arena: agents building gamesNot measured1130Only Kimi K2.7 Code measured
Design Arena: HTML slidesNot measured1201Only Kimi K2.7 Code measured
Design Arena: PowerPoint slidesNot measured1154Only Kimi K2.7 Code measured
Finance Agent (Vals AI)35.8%Not measuredOnly Kimi K2.5 measured
Legal Research Bench (Vals AI)15.9%Not measuredOnly Kimi K2.5 measured
Excel Modeling Benchmark (Vals AI)28.5%Not measuredOnly Kimi K2.5 measured
τ²-Bench Telecom95.9%90.1%Kimi K2.5
APEX-AgentsNot measured37.6%Only Kimi K2.7 Code measured
OSWorld63.3%Not measuredOnly Kimi K2.5 measured
Vending-Bench 2$1,198$5,083Kimi K2.7 Code
Neither model has a result on the other 23 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index03
Reasoning benchmarks for Kimi K2.5 and Kimi K2.7 Code
BenchmarkKimi K2.5Kimi K2.7 CodeLeader
GPQA Diamond84.1%Not measuredOnly Kimi K2.5 measured
MMMU Pro (Vals AI)84.3%Not measuredOnly Kimi K2.5 measured
GPQA Diamond (Artificial Analysis run)87.9%89.6%Kimi K2.7 Code
Humanity's Last Exam (Artificial Analysis run)30.7%35.0%Kimi K2.7 Code
GPQA Diamond (Epoch AI run)Not measured87.9%Only Kimi K2.7 Code measured
ARC-AGI65.3%Not measuredOnly Kimi K2.5 measured
ARC-AGI-211.8%Not measuredOnly Kimi K2.5 measured
SimpleBench46.8%57.9%Kimi K2.7 Code
Chess puzzlesNot measured14.0%Only Kimi K2.7 Code measured
LMArena Text: hard prompts1453Not measuredOnly Kimi K2.5 measured
LMArena Text: expert1466Not measuredOnly Kimi K2.5 measured
LMArena Text: hard prompts in English1460Not measuredOnly Kimi K2.5 measured
LMArena Vision: diagrams1282Not measuredOnly Kimi K2.5 measured
LMArena Vision: homework1297Not measuredOnly Kimi K2.5 measured
Neither model has a result on the other 7 boards here
MathDirectional onlyPublic boards, shown as published, kept out of the index6 boards, none shared
Math benchmarks for Kimi K2.5 and Kimi K2.7 Code
BenchmarkKimi K2.5Kimi K2.7 CodeLeader
AIME95.6%Not measuredOnly Kimi K2.5 measured
OTIS Mock AIMENot measured95.6%Only Kimi K2.7 Code measured
FrontierMath, tiers 1 to 3Not measured54.0%Only Kimi K2.7 Code measured
FrontierMath, tier 4Not measured12.2%Only Kimi K2.7 Code measured
LMArena Text: math1470Not measuredOnly Kimi K2.5 measured
LMArena Text: mathematical work1473Not measuredOnly Kimi K2.5 measured
Neither model has a result on the other 7 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index01
Knowledge benchmarks for Kimi K2.5 and Kimi K2.7 Code
BenchmarkKimi K2.5Kimi K2.7 CodeLeader
MMLU Pro85.9%Not measuredOnly Kimi K2.5 measured
CaseLaw (Vals AI)58.7%Not measuredOnly Kimi K2.5 measured
CorpFin (Vals AI)68.3%Not measuredOnly Kimi K2.5 measured
TaxEval (Vals AI)74.2%Not measuredOnly Kimi K2.5 measured
MortgageTax (Vals AI)66.5%Not measuredOnly Kimi K2.5 measured
MedQA (Vals AI)94.4%Not measuredOnly Kimi K2.5 measured
MedCode (Vals AI)39.3%Not measuredOnly Kimi K2.5 measured
SimpleQA Verified34.3%36.5%Kimi K2.7 Code
LMArena Text: science1459Not measuredOnly Kimi K2.5 measured
LMArena Text: medicine and healthcare1448Not measuredOnly Kimi K2.5 measured
LMArena Text: legal and government1441Not measuredOnly Kimi K2.5 measured
LMArena Text: business and finance1427Not measuredOnly Kimi K2.5 measured
Neither model has a result on the other 5 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index08
Web design benchmarks for Kimi K2.5 and Kimi K2.7 Code
BenchmarkKimi K2.5Kimi K2.7 CodeLeader
Design Arena: websites12491264Kimi K2.7 Code
Design Arena: UI components12361259Kimi K2.7 Code
Design Arena: game dev12071220Kimi K2.7 Code
Design Arena: data viz12191226Kimi K2.7 Code
Design Arena: 3D12051241Kimi K2.7 Code
Design Arena: all code categories12361253Kimi K2.7 Code
Design Arena: SVG11491176Kimi K2.7 Code
Design Arena: ASCII art11651196Kimi K2.7 Code
Neither model has a result on the other 1 board here
GeneralLike-for-likePublic boards, shown as published, kept out of the index11
General benchmarks for Kimi K2.5 and Kimi K2.7 Code
BenchmarkKimi K2.5Kimi K2.7 CodeLeader
SAGE (Vals AI)49.9%Not measuredOnly Kimi K2.5 measured
MedScribe (Vals AI)76.4%Not measuredOnly Kimi K2.5 measured
IFBench70.2%63.1%Kimi K2.5
Long context reasoning (AA-LCR)78.0%79.3%Kimi K2.7 Code
Fiction.LiveBench86.1%Not measuredOnly Kimi K2.5 measured
CL-bench19.3%Not measuredOnly Kimi K2.5 measured
Surface Evolver BenchNot measured48.8%Only Kimi K2.7 Code measured
CL-bench Life13.2%Not measuredOnly Kimi K2.5 measured
LMArena Text1445Not measuredOnly Kimi K2.5 measured
LMArena Text: instruction following1431Not measuredOnly Kimi K2.5 measured
LMArena Text: creative writing1422Not measuredOnly Kimi K2.5 measured
LMArena Vision1269Not measuredOnly Kimi K2.5 measured
LMArena Text: English1453Not measuredOnly Kimi K2.5 measured
LMArena Text: non-English1433Not measuredOnly Kimi K2.5 measured
LMArena Text: Chinese1494Not measuredOnly Kimi K2.5 measured
LMArena Text: German1441Not measuredOnly Kimi K2.5 measured
LMArena Text: Japanese1423Not measuredOnly Kimi K2.5 measured
LMArena Text: Korean1412Not measuredOnly Kimi K2.5 measured
LMArena Text: French1454Not measuredOnly Kimi K2.5 measured
LMArena Text: Spanish1450Not measuredOnly Kimi K2.5 measured
LMArena Text: Russian1435Not measuredOnly Kimi K2.5 measured
LMArena Text: Polish1445Not measuredOnly Kimi K2.5 measured
LMArena Text: multi-turn1444Not measuredOnly Kimi K2.5 measured
LMArena Text: longer queries1445Not measuredOnly Kimi K2.5 measured
LMArena Text: writing and language1425Not measuredOnly Kimi K2.5 measured
LMArena Text: entertainment and media1417Not measuredOnly Kimi K2.5 measured
LMArena Vision: English1274Not measuredOnly Kimi K2.5 measured
LMArena Vision: Chinese1328Not measuredOnly Kimi K2.5 measured
LMArena Vision: reading text1276Not measuredOnly Kimi K2.5 measured
LMArena Vision: humor1284Not measuredOnly Kimi K2.5 measured
LMArena Vision: captioning1257Not measuredOnly Kimi K2.5 measured
LMArena Vision: recognizing things1285Not measuredOnly Kimi K2.5 measured
LMArena Vision: creative writing1269Not measuredOnly Kimi K2.5 measured
Neither model has a result on the other 13 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for Kimi K2.5 and Kimi K2.7 Code
BenchmarkKimi K2.5Kimi K2.7 CodeLeader
Neither model has a result on 5 boards here

What each workload costs

Three fixed token mixes at each model's listed price, so per-token rates become a bill.

Chat turn

1K input + 500 output tokens

Kimi K2.5$0.0018Fits in one request
Kimi K2.7 Code$0.0023Fits in one request

Kimi K2.5 costs less

Repository review

50K input + 3K output tokens

Kimi K2.5$0.033Fits in one request
Kimi K2.7 Code$0.044Fits in one request

Kimi K2.5 costs less

Agent loop

200K input + 10K output tokens

Kimi K2.5$0.125Fits in one request
Kimi K2.7 Code$0.168Fits in one request

Kimi K2.5 costs less

Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.

Specification differences

The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.

Specifications of Kimi K2.5 and Kimi K2.7 Code
SpecificationMoonshotAI Kimi K2.5MoonshotAI Kimi K2.7 Code
MakerMoonshot AIMoonshot AI
WeightsOpen weightsOpen weights
ReleasedJanuary 27, 2026June 12, 2026
Context window262K262K
Input $/Mtok$0.50$0.6712
Output $/Mtok$2.50$3.35
Coding index58.539.1
Independent measurements behind it21
Cost per solved task$0.24$0.35
SourceOpenRouter listingOpenRouter listing

Questions

Is Kimi K2.5 better than Kimi K2.7 Code for coding?

Pick Kimi K2.5 for coding: it scores 58.5 on the Vibe Coding Index against 39.1 for Kimi K2.7 Code. Pick Kimi K2.7 Code when what matters is LMArena WebDev or AA Coding Index, where it leads.

Which is cheaper, Kimi K2.5 or Kimi K2.7 Code?

Kimi K2.5 costs $0.50 per million input tokens and $2.50 per million output tokens; Kimi K2.7 Code costs $0.6712 and $3.35. Per solved SWE-bench Verified task, Kimi K2.5 costs $0.24 and Kimi K2.7 Code $0.35.

Which has the bigger context window, Kimi K2.5 or Kimi K2.7 Code?

Neither: both take 262K tokens.

How many benchmarks is the Kimi K2.5 vs Kimi K2.7 Code comparison based on?

36 benchmarks have a result for both models as of October 21, 2026. Kimi K2.5's coding index rests on 2 independent benchmarks and Kimi K2.7 Code's on 1. Where only one model has a result, the page says so instead of naming a winner.

Compare next