LLMBenchmark.ai

Enter opens the highlighted result. Esc closes.

GPT-5.4 vs Kimi K2.6

Updated October 21, 2026. 94 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.

Replace GPT-5.4 Replace Kimi K2.6
GPT-5.4OpenAI
Coding
37.7
Overall
39.0
Output $/M
$15.00
68
Kimi K2.6Moonshot AI
Coding
44.2
Overall
27.0
Output $/M
$3.41
26
vs boards won of 94 shared

Decision reading

Pick Kimi K2.6 for coding: it scores 44.2 on the Vibe Coding Index against 37.7 for GPT-5.4.

Pick GPT-5.4 when the context window is larger, 1.05M against 262K tokens, or when what matters is AA Coding Index or SWE-bench Verified (Vals AI), where it leads.

Model AGPT-5.4OpenAI37.7/1001 independent benchmark$1.02 per solved taskModel BKimi K2.6Moonshot AI44.2/1001 independent benchmark$0.35 per solved task
Shared results
94
GPT-5.4 only
15
Kimi K2.6 only
15
Like-for-like categories
8 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
GPT-5.4 37.7 from 1 benchmark, Kimi K2.6 44.2 from 1 benchmark.
Widest category
Math, 15.3 points apart on average over 6 shared benchmarks.
Widest single gap
ProofBench: GPT-5.4 56.0% against Kimi K2.6 16.0%, 40.0 points apart on a 0 to 100 scale.
Output price
GPT-5.4 $15.00 against Kimi K2.6 $3.41 per million output tokens, 4.4x apart.
Context window
GPT-5.4 takes 1.05M tokens, Kimi K2.6 262K.

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
CategoryGPT-5.4BarsKimi K2.6Ties
Coding105
Intelligence20
Agentic work16
Reasoning130
Math60
Knowledge93
Web design17
General265
All shared boards6826
CodingLike-for-likeFour of these boards feed the Vibe Coding Index105
Coding benchmarks for GPT-5.4 and Kimi K2.6
BenchmarkGPT-5.4Kimi K2.6Leader
LMArena WebDev14641509Kimi K2.6
AA Coding Index71.161.8GPT-5.4
LiveCodeBench (Vals AI)84.1%86.8%Kimi K2.6
SWE-bench Verified (Vals AI)78.2%76.2%GPT-5.4
Vibe Code Bench (Vals AI)67.4%37.9%GPT-5.4
Code migration (Vals AI)35.0%27.8%GPT-5.4
ProgramBench (Vals AI)0.5%0.0%GPT-5.4
Terminal-Bench 2.1 (Vals AI)Not measured53.6%Only Kimi K2.6 measured
SkillsBench (Vals AI)51.7%Not measuredOnly GPT-5.4 measured
SciCodeNot measured51.5%Only Kimi K2.6 measured
Terminal-Bench Hard (Artificial Analysis run)57.6%43.9%GPT-5.4
Terminal-Bench81.8%Not measuredOnly GPT-5.4 measured
GSO-Bench31.4%Not measuredOnly GPT-5.4 measured
WeirdML77.7%55.9%GPT-5.4
MirrorCode15.6%Not measuredOnly GPT-5.4 measured
ALE-Bench1607.01092.7GPT-5.4
LMArena Text: coding14961489GPT-5.4
LMArena Text: software and IT14911484GPT-5.4
LMArena WebDev: React14541505Kimi K2.6
LMArena WebDev: HTML14781485Kimi K2.6
LMArena WebDev: from an image14361521Kimi K2.6
Neither model has a result on the other 17 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
Intelligence benchmarks for GPT-5.4 and Kimi K2.6
BenchmarkGPT-5.4Kimi K2.6Leader
AA Intelligence Index39.027.0GPT-5.4
Epoch Capabilities Index157.3150.9GPT-5.4
Agentic workLike-for-likePublic boards, shown as published, kept out of the index16
Agentic work benchmarks for GPT-5.4 and Kimi K2.6
BenchmarkGPT-5.4Kimi K2.6Leader
AA Agentic IndexNot measured20.5Only Kimi K2.6 measured
Design Arena: agents building web apps10591268Kimi K2.6
Design Arena: agents building full-stack apps10171155Kimi K2.6
Design Arena: agents building mobile apps10871156Kimi K2.6
Design Arena: agents building Android apps10371166Kimi K2.6
Design Arena: agents building Godot games11351120GPT-5.4
Design Arena: agents building gamesNot measured1141Only Kimi K2.6 measured
Design Arena: HTML slidesNot measured1184Only Kimi K2.6 measured
Design Arena: PowerPoint slidesNot measured1128Only Kimi K2.6 measured
Design Arena: spreadsheetsNot measured1276Only Kimi K2.6 measured
Finance Agent (Vals AI)Not measured44.9%Only Kimi K2.6 measured
Tax Agent Bench (Vals AI)Not measured12.4%Only Kimi K2.6 measured
Legal Research Bench (Vals AI)Not measured15.9%Only Kimi K2.6 measured
Excel Modeling Benchmark (Vals AI)Not measured57.9%Only Kimi K2.6 measured
τ²-Bench Telecom87.1%95.9%Kimi K2.6
APEX-Agents52.4%Not measuredOnly GPT-5.4 measured
METR time horizons74.3%Not measuredOnly GPT-5.4 measured
OSWorld 2.0Not measured4.6%Only Kimi K2.6 measured
DeepResearch Bench35.1%Not measuredOnly GPT-5.4 measured
PostTrainBench19.0%Not measuredOnly GPT-5.4 measured
Vending-Bench 2$6,144$6,205Kimi K2.6
Neither model has a result on the other 18 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index130
Reasoning benchmarks for GPT-5.4 and Kimi K2.6
BenchmarkGPT-5.4Kimi K2.6Leader
GPQA Diamond91.7%89.1%GPT-5.4
MMMU Pro (Vals AI)87.5%86.3%GPT-5.4
GPQA Diamond (Artificial Analysis run)92.0%91.1%GPT-5.4
Humanity's Last Exam (Artificial Analysis run)43.7%37.5%GPT-5.4
GPQA Diamond (Epoch AI run)93.3%90.8%GPT-5.4
ARC-AGI93.7%Not measuredOnly GPT-5.4 measured
ARC-AGI-274.0%Not measuredOnly GPT-5.4 measured
Chess puzzles48.0%20.0%GPT-5.4
Mystery game puzzles37.0%18.0%GPT-5.4
EBR-bench25.4%2.4%GPT-5.4
LMArena Text: hard prompts14851470GPT-5.4
LMArena Text: expert15071491GPT-5.4
LMArena Text: hard prompts in English14791471GPT-5.4
LMArena Vision: diagrams13271298GPT-5.4
LMArena Vision: homework13361310GPT-5.4
Neither model has a result on the other 6 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index60
Math benchmarks for GPT-5.4 and Kimi K2.6
BenchmarkGPT-5.4Kimi K2.6Leader
AIME96.7%Not measuredOnly GPT-5.4 measured
OTIS Mock AIME97.8%96.1%GPT-5.4
FrontierMath, tiers 1 to 378.6%57.2%GPT-5.4
FrontierMath, tier 449.0%25.6%GPT-5.4
ProofBench56.0%16.0%GPT-5.4
LMArena Text: math14881474GPT-5.4
LMArena Text: mathematical work14971480GPT-5.4
Neither model has a result on the other 6 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index93
Knowledge benchmarks for GPT-5.4 and Kimi K2.6
BenchmarkGPT-5.4Kimi K2.6Leader
MMLU Pro87.5%87.6%Kimi K2.6
LegalBench (Vals AI)86.0%84.7%GPT-5.4
CaseLaw (Vals AI)63.8%61.2%GPT-5.4
CorpFin (Vals AI)65.3%66.7%Kimi K2.6
TaxEval (Vals AI)74.0%74.7%Kimi K2.6
MortgageTax (Vals AI)68.3%65.8%GPT-5.4
MedQA (Vals AI)96.1%Not measuredOnly GPT-5.4 measured
MedCode (Vals AI)41.3%40.1%GPT-5.4
SimpleQA Verified45.1%34.9%GPT-5.4
LMArena Text: science14791468GPT-5.4
LMArena Text: medicine and healthcare14581457GPT-5.4
LMArena Text: legal and government14861462GPT-5.4
LMArena Text: business and finance14741445GPT-5.4
Neither model has a result on the other 4 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index17
Web design benchmarks for GPT-5.4 and Kimi K2.6
BenchmarkGPT-5.4Kimi K2.6Leader
Design Arena: websites11991269Kimi K2.6
Design Arena: UI components12151266Kimi K2.6
Design Arena: game dev11951251Kimi K2.6
Design Arena: data viz11961249Kimi K2.6
Design Arena: 3D10651268Kimi K2.6
Design Arena: all code categories11801272Kimi K2.6
Design Arena: SVG11411181Kimi K2.6
Design Arena: ASCII art11881164GPT-5.4
Neither model has a result on the other 1 board here
GeneralLike-for-likePublic boards, shown as published, kept out of the index265
General benchmarks for GPT-5.4 and Kimi K2.6
BenchmarkGPT-5.4Kimi K2.6Leader
Public Benefits Bench (Vals AI)Not measured56.6%Only Kimi K2.6 measured
SAGE (Vals AI)43.3%50.2%Kimi K2.6
MedScribe (Vals AI)77.5%78.2%Kimi K2.6
Vals Multimodal IndexNot measured56.4%Only Kimi K2.6 measured
IFBench73.9%76.0%Kimi K2.6
Long context reasoning (AA-LCR)82.0%81.0%GPT-5.4
DTBench94.4%90.9%GPT-5.4
LMCA52.0%37.3%GPT-5.4
CL-bench27.9%Not measuredOnly GPT-5.4 measured
Furniture assembly37.5%21.7%GPT-5.4
ExploitBenchNot measured18.4%Only Kimi K2.6 measured
CL-bench Life21.7%Not measuredOnly GPT-5.4 measured
LMArena Text14691455GPT-5.4
LMArena Text: instruction following14691451GPT-5.4
LMArena Text: creative writing14391434GPT-5.4
LMArena Vision13031282GPT-5.4
LMArena Text: English14671460GPT-5.4
LMArena Text: non-English14651446GPT-5.4
LMArena Text: Chinese15201520Kimi K2.6
LMArena Text: German14721451GPT-5.4
LMArena Text: Japanese14851443GPT-5.4
LMArena Text: Korean14491427GPT-5.4
LMArena Text: French14931471GPT-5.4
LMArena Text: Spanish14521464Kimi K2.6
LMArena Text: Russian14801446GPT-5.4
LMArena Text: Polish14721438GPT-5.4
LMArena Text: multi-turn14821453GPT-5.4
LMArena Text: longer queries14731468GPT-5.4
LMArena Text: writing and language14581439GPT-5.4
LMArena Text: entertainment and media14381428GPT-5.4
LMArena Vision: English13011287GPT-5.4
LMArena Vision: Chinese13541346GPT-5.4
LMArena Vision: reading text13121291GPT-5.4
LMArena Vision: humor12841280GPT-5.4
LMArena Vision: recognizing things1251Not measuredOnly GPT-5.4 measured
LMArena Vision: creative writing12831273GPT-5.4
LMArena Document14711451GPT-5.4
Neither model has a result on the other 9 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for GPT-5.4 and Kimi K2.6
BenchmarkGPT-5.4Kimi K2.6Leader
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

GPT-5.4$0.010Fits in one request
Kimi K2.6$0.0024Fits in one request

Kimi K2.6 costs less

Repository review

50K input + 3K output tokens

GPT-5.4$0.170Fits in one request
Kimi K2.6$0.043Fits in one request

Kimi K2.6 costs less

Agent loop

200K input + 10K output tokens

GPT-5.4$0.650Fits in one request
Kimi K2.6$0.164Fits in one request

Kimi K2.6 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 GPT-5.4 and Kimi K2.6
SpecificationOpenAI GPT-5.4MoonshotAI Kimi K2.6
MakerOpenAIMoonshot AI
WeightsNo public weights listedOpen weights
ReleasedMarch 5, 2026April 20, 2026
Context window1.05M262K
Input $/Mtok$2.50$0.65
Output $/Mtok$15.00$3.41
Coding index37.744.2
Independent measurements behind it11
Cost per solved task$1.02$0.35
SourceOpenRouter listingOpenRouter listing

Questions

Is GPT-5.4 better than Kimi K2.6 for coding?

Pick Kimi K2.6 for coding: it scores 44.2 on the Vibe Coding Index against 37.7 for GPT-5.4. Pick GPT-5.4 when the context window is larger, 1.05M against 262K tokens, or when what matters is AA Coding Index or SWE-bench Verified (Vals AI), where it leads.

Which is cheaper, GPT-5.4 or Kimi K2.6?

GPT-5.4 costs $2.50 per million input tokens and $15.00 per million output tokens; Kimi K2.6 costs $0.65 and $3.41. Per solved SWE-bench Verified task, GPT-5.4 costs $1.02 and Kimi K2.6 $0.35.

Which has the bigger context window, GPT-5.4 or Kimi K2.6?

GPT-5.4, with 1.05M tokens against 262K.

How many benchmarks is the GPT-5.4 vs Kimi K2.6 comparison based on?

94 benchmarks have a result for both models as of October 21, 2026. GPT-5.4's coding index rests on 1 independent benchmark and Kimi K2.6's on 1. Where only one model has a result, the page says so instead of naming a winner.

Compare next