Skip to main content

OpenAIMoonshot AILLM matchup

GPT-5.4vs.Kimi K3

Which one is better?

Kimi K3 performs better than GPT-5.4, but costs 1.2 times as much for input and about the same for output.

Kimi K3 ahead
IntelligenceOverall index, 100 points = average
86.4

GPT-5.4

Rank 50

101.5

Kimi K3

Rank 26

about as fast
SpeedTokens per second
64

GPT-5.4

66

Kimi K3

about the same price
Price$ per 1M tokens
$5.63

GPT-5.4

$6

Kimi K3

Popularity0 to 100
n/a

GPT-5.4

85

Kimi K3

Rank 6 of 508

Above 272,000 tokens of context, GPT-5.4 costs more: $5 input and $22.50 output per 1M tokens.

No clear advantage for GPT-5.4.

Choose Kimi K3 for

LegalHealthFinanceCodingAgentsReasoning (not certain)

Where is the difference?

Strengths at a glance

GPT-5.4Kimi K3Median of all ranked models

859510511575CodingKimi K3Reasoningslightly KimiK3AgentsKimi K3FinanceKimi K3LegalKimi K3HealthKimi K3
CodingKimi K3 ahead
GPT-5.485.0Rank 60
Kimi K397.3Rank 43
Median of ranked models87.5

95 % range: GPT-5.4 75.3 to 94.6, Kimi K3 89.9 to 104.6.

AreaGPT-5.4Kimi K3

The same values and verdicts as the bars below. In the close-up the net starts at 75 points, just below the weaker model's lowest value; the reference group average is 100. Not in the net because a model has no own value there or the data are too thin: Mathematics, Knowledge, General, Multimodal.

Kimi K3 is ahead in 5 of 10 areas, GPT-5.4 in none.

too little data: 4Kimi K3 slightly ahead: 1Kimi K3 ahead: 5

Areas in the overall index
CodingKimi K3 +1285.0 to 97.3
AgentsKimi K3 +790.5 to 97.4
Reasoningleaning Kimi K3not certain · 3 and 5 tests
Generaltoo little data86.1 to 106.0
Multimodaltoo little data88.8 to 101.6
Knowledgetoo little data87.0 to 95.3
Mathematicsonly Kimi K3 has a score
Specialist fields
LegalKimi K3 +2683.9 to 109.8
HealthKimi K3 +2580.4 to 105.0
FinanceKimi K3 +1985.9 to 105.3

Bar length shows the gap in index points (full width 15 points, one standard deviation of the reference group). Solid means clearly ahead, hatched not certain yet, a thin grey bar level.

Show 3 more areas that do not count toward the overall index
Cybersecurityonly Kimi K3 has a score
Productivityonly Kimi K3 has a score
Educationonly Kimi K3 has a score

Bar length shows the gap in index points (full width 15 points, one standard deviation of the reference group). Solid means clearly ahead, hatched not certain yet, a thin grey bar level.

Which model for which task?

Everyday tasks, translated from the area values above. Each row names the values it rests on.

Writing and debugging codeKimi K3Coding 85.0 to 97.3
Maths and data analysisleaning Kimi K3Mathematics without a score for GPT-5.4, Reasoning 87.9 to 97.0
Facts and researchopenKnowledge 87.0 to 95.3 (too little data)
Agents and workflowsKimi K3Agents 90.5 to 97.4
Fast chat repliesabout the same64 to 66 tokens per second
Security analysisopenCybersecurity without a score for GPT-5.4
Many requests on a small budgetabout the same$5.63 to $6 per 1M tokens
Finance, legal and medical textsKimi K3Finance 85.9 to 105.3, Legal 83.9 to 109.8, Health 80.4 to 105.0
Images, screenshots and PDFsopenMultimodal 88.8 to 101.6 (too little data)

GPT-5.4 vs. Kimi K3 in our own test

The same prompt for both models. We show the latest attempt without fixes; it does not count toward the overall index.

Pelican on a bicycle

Prompt

Generate an SVG of a pelican riding a bicycle

Task origin: Simon Willison

GPT-5.4

GPT-5.4: Pelican on a bicycle
A community classic for free-form SVG drawing.

Oct 1, 2026 · 117 s · $0.11

Test details
Run date
Oct 1, 2026
API provider
openrouter
Requested API model
openai/gpt-5.4
Reasoning setting
high
Token limit including reasoning
8,192 tokens
Output tokens
7,339 tokens
Cost of the attempt
$0.1101
Response time
116.8 s
API model reported in response
openai/gpt-5.4
Pinned endpoint
OpenAI | openai/gpt-5.4-20260305
Temperature
Provider default
Test protocol
community-visual-tests-v1
Technical checks
Passed, not a quality rating

Visually reviewed. This individual test does not establish overall model quality.

Earlier attempts (1)

  • Sep 7, 2026API request failedReasoning setting: high

Kimi K3

Kimi K3: Pelican on a bicycle
A community classic for free-form SVG drawing.

Oct 1, 2026 · 173 s · $0.08

Test details
Run date
Oct 1, 2026
API provider
openrouter
Requested API model
moonshotai/kimi-k3
Reasoning setting
high
Token limit including reasoning
8,192 tokens
Output tokens
5,145 tokens
Cost of the attempt
$0.0775
Response time
173.4 s
API model reported in response
moonshotai/kimi-k3
Pinned endpoint
Moonshot AI | moonshotai/kimi-k3-20260715
Temperature
Provider default
Test protocol
community-visual-tests-v1
Technical checks
Passed, not a quality rating

Visually reviewed. This individual test does not establish overall model quality.

Earlier attempts (1)

  • Sep 7, 2026API request failedReasoning setting: high
SVG short test

SVG short test

One SVG task with the same fixed settings for every model, without hidden thinking.

Prompt

Return only one standalone SVG of a pelican riding a bicycle. Use viewBox="0 0 800 600", simple shapes, two clearly visible wheels and a recognizable pelican with a long beak. At most 30 SVG elements and 1800 characters. No Markdown, explanations, comments, scripts, animation, CSS or external resources.

GPT-5.4

Unavailable in this round

Kimi K3

Kimi K3: SVG short test

Oct 14, 2026 · $0.0049

Test details
Test run
Oct 14, 2026
Provider
chutes/mxfp4
API model
moonshotai/kimi-k3
Output
495 tokens
Reasoning
API reports 0 reasoning tokens
Cost of the attempt
$0.0049
Test round
43b8c5c46830

GPT-5.4 and Kimi K3 in the overall index

Every model with a place in the overall index, both models of this matchup highlighted. You can add or hide other models.

Overall index (Gradually Meta-Score v6)

20 of 95 models

Claude Fable 5.2max130.9
Gemini 4 Carbonhigh127.0
Claude Opus 5.5max125.7
Claude Fable 5.1max124.1
Claude Sonnet 5.5max121.9
Gemini 4 Argonhigh121.2
GPT-6 Astramax119.6
Muse Wonder Fullmax118.4
GPT-6.1 Solmax117.5
Claude Opus 5max117.3
Muse Wondermax113.7
Claude Fable 5max112.8
Kimi K3.1 Promax112.6
Muse Ember 1.0max110.5
GPT-5.6 Solmax109.3
GPT-6 Solmax108.4
Gemini 3.7 Flashhigh107.7
DeepSeek V4.1 Flash 1017max107.1
Kimi K3max101.5
GPT-5.4xhigh86.4

Popularity compared

How often the models are searched for, used and mentioned, on a scale from 0 to 100.

GPT-5.4 has no popularity score yet.

Over time

Weekly values from October 18, 2026 to November 16, 2026

GPT-5.4Kimi K3

All data side by side

The better value is bold. For the overall index only when the gap is clear.

MetricGPT-5.4Kimi K3
Intelligence
Overall indexpoints, 100 = reference group average, higher is better86.4101.5
RankRank 50 of 95Rank 26 of 95
Price per 1M tokens
Inputwhat you send$2.50above 272,000 tokens: $5$3
Outputwhat the model writes$15above 272,000 tokens: $22.50$15
Cost per task (own tests)average of 3 original-protocol tasks both models passed (SVG short test not included)$0.19$0.055
Speed
Tokens per second6466
Popularity
Popularitysearch, usage and mentions, 0 to 100n/a85Rank 6 of 508
Technical
Context windowhow much text the model reads at once1.05 M tokens≈ 1,580 A4 pages1.05 M tokens≈ 1,570 A4 pages
Knowledge cutoffAug 31, 2025n/a
ReleasedMar 5, 2026Jul 16, 2026
AvailabilityActiveActive
Open weightsnoyes

Prices without caching or batch discounts · Page count as a rule of thumb from Artificial Analysis

Is there a better choice?

Successor of GPT-5.4GPT-5.5

Overall index 101.4, rank 27$5 input and $30 output per 1M tokens

Kimi K3 vs. GPT-5.5
Successor of Kimi K3Kimi K3.1 Pro

Overall index 112.6, rank 13$2 input and $10 output per 1M tokens

GPT-5.4 vs. Kimi K3.1 Pro
Cheaper than GPT-5.4Gemini 4 Carbon

Overall index 127.0, rank 2$2 input and $10 output per 1M tokens

Kimi K3 vs. Gemini 4 Carbon

Shared benchmarks

Single results from the same benchmark version, task and metric. A value is bold only with the same test setup and a clear gap.

A: GPT-5.4B: Kimi K3

Code MigrationOverall · higher is betterTest setup not documented as identicalGPT-5.4: 35.0%Kimi K3: 16.1%Vals AIAs of 10/29/2026
CorpFin v2Overall · higher is betterTest setup not documented as identicalGPT-5.4: 65.3%Kimi K3: 71.6%Vals AIAs of 08/12/2026
GPQA DiamondOverall · higher is betterTest setup not documented as identicalGPT-5.4: 91.7%Kimi K3: 92.9%Vals AIAs of 09/01/2026
LegalBenchOverall · higher is betterTest setup not documented as identicalGPT-5.4: 86.0%Kimi K3: 86.2%Vals AIAs of 10/13/2026
LiveCodeBenchOverall · higher is betterTest setup not documented as identicalGPT-5.4: 84.1%Kimi K3: 87.2%Vals AIAs of 09/01/2026
LMArena Code Arena WebDevOverall · higher is betterTest setup not documented as identicalGPT-5.4: 1,400Kimi K3: 1,658LMArenaAs of 11/15/2026
LMArena Text, Style ControlOverall · higher is betterTest setup not documented as identicalGPT-5.4: 1,475Kimi K3: 1,487LMArenaAs of 11/11/2026
MedCodeOverall · higher is betterTest setup not documented as identicalGPT-5.4: 41.3%Kimi K3: 49.4%Vals AIAs of 10/29/2026
MedScribeOverall · higher is betterTest setup not documented as identicalGPT-5.4: 77.5%Kimi K3: 88.0%Vals AIAs of 10/29/2026
MMLU ProOverall · higher is betterTest setup not documented as identicalGPT-5.4: 87.5%Kimi K3: 88.0%Vals AIAs of 09/01/2026
Show 54 more benchmarks
MMMU ProOverall · higher is betterTest setup not documented as identicalGPT-5.4: 87.5%Kimi K3: 88.2%Vals AIAs of 09/01/2026
MMMU-ProOverall · higher is betterGPT-5.4: 78.0%Kimi K3: 78.5%Artificial AnalysisAs of 11/07/2026
MortgageTaxOverall · higher is betterTest setup not documented as identicalGPT-5.4: 68.3%Kimi K3: 66.3%Vals AIAs of 09/01/2026
ProgramBenchOverall · higher is betterTest setup not documented as identicalGPT-5.4: 0%Kimi K3: 2%Vals AIAs of 11/05/2026
SWE-bench VerifiedOverall · higher is betterTest setup not documented as identicalGPT-5.4: 78.2%Kimi K3: 93.4%Vals AIAs of 09/01/2026
TaxEval v2Overall · higher is betterTest setup not documented as identicalGPT-5.4: 74.0%Kimi K3: 75.7%Vals AIAs of 09/01/2026
Vibe Code Bench v1.1Overall · higher is betterTest setup not documented as identicalGPT-5.4: 67.4%Kimi K3: 85.0%Vals AIAs of 11/05/2026
Code MigrationCLI · higher is betterTest setup not documented as identicalGPT-5.4: 23.8%Kimi K3: 1.5%Vals AIAs of 10/29/2026
CorpFin v2Exact Pages · higher is betterTest setup not documented as identicalGPT-5.4: 64.1%Kimi K3: 69.3%Vals AIAs of 08/12/2026
GPQA DiamondFew-Shot CoT · higher is betterTest setup not documented as identicalGPT-5.4: 91.4%Kimi K3: 93.4%Vals AIAs of 09/01/2026
LegalBenchConclusion Tasks · higher is betterTest setup not documented as identicalGPT-5.4: 88.9%Kimi K3: 90.1%Vals AIAs of 10/13/2026
LiveCodeBenchEasy · higher is betterTest setup not documented as identicalGPT-5.4: 99.1%Kimi K3: 98.8%Vals AIAs of 09/01/2026
LMArena Code Arena WebDevHTML development · higher is betterTest setup not documented as identicalGPT-5.4: 1,423Kimi K3: 1,640LMArenaAs of 11/15/2026
LMArena Text, Style ControlBusiness, management and finance · higher is betterTest setup not documented as identicalGPT-5.4: 1,482Kimi K3: 1,488LMArenaAs of 11/11/2026
MMLU ProBiology · higher is betterTest setup not documented as identicalGPT-5.4: 92.9%Kimi K3: 93.9%Vals AIAs of 09/01/2026
MortgageTaxNumerical Extraction · higher is betterTest setup not documented as identicalGPT-5.4: 76.0%Kimi K3: 78.9%Vals AIAs of 09/01/2026
ProgramBenchAlmost Resolved · higher is betterTest setup not documented as identicalGPT-5.4: 7.0%Kimi K3: 17.5%Vals AIAs of 11/05/2026
SWE-bench Verified<15 min fix · higher is betterTest setup not documented as identicalGPT-5.4: 88.1%Kimi K3: 92.3%Vals AIAs of 09/01/2026
TaxEval v2Correctness · higher is betterTest setup not documented as identicalGPT-5.4: 64.9%Kimi K3: 65.3%Vals AIAs of 09/01/2026
Code MigrationCOBOL · higher is betterTest setup not documented as identicalGPT-5.4: 68.6%Kimi K3: 60.0%Vals AIAs of 10/29/2026
CorpFin v2Max Fitting Context · higher is betterTest setup not documented as identicalGPT-5.4: 64.9%Kimi K3: 72.7%Vals AIAs of 08/12/2026
GPQA DiamondZero-Shot CoT · higher is betterTest setup not documented as identicalGPT-5.4: 91.9%Kimi K3: 92.4%Vals AIAs of 09/01/2026
LegalBenchInterpretation Tasks · higher is betterTest setup not documented as identicalGPT-5.4: 83.9%Kimi K3: 84.1%Vals AIAs of 10/13/2026
LiveCodeBenchHard · higher is betterTest setup not documented as identicalGPT-5.4: 67.7%Kimi K3: 74.3%Vals AIAs of 09/01/2026
LMArena Code Arena WebDevImage to WebDev · higher is betterTest setup not documented as identicalGPT-5.4: 1,441Kimi K3: 1,585LMArenaAs of 11/15/2026
LMArena Text, Style ControlChinese · higher is betterTest setup not documented as identicalGPT-5.4: 1,505Kimi K3: 1,540LMArenaAs of 11/11/2026
MMLU ProBusiness · higher is betterTest setup not documented as identicalGPT-5.4: 91.8%Kimi K3: 92.0%Vals AIAs of 09/01/2026
MortgageTaxSemantic Extraction · higher is betterTest setup not documented as identicalGPT-5.4: 60.7%Kimi K3: 53.8%Vals AIAs of 09/01/2026
ProgramBenchFully Resolved · higher is betterTest setup not documented as identicalGPT-5.4: 0%Kimi K3: 2%Vals AIAs of 11/05/2026
SWE-bench Verified>4 hours · higher is betterTest setup not documented as identicalGPT-5.4: 0.0%Kimi K3: 66.7%Vals AIAs of 09/01/2026
TaxEval v2Stepwise Reasoning · higher is betterTest setup not documented as identicalGPT-5.4: 83.0%Kimi K3: 86.1%Vals AIAs of 09/01/2026
Code MigrationCode Quality · higher is betterTest setup not documented as identicalGPT-5.4: 3.632 / 5Kimi K3: 3.899 / 5Vals AIAs of 10/29/2026
CorpFin v2Shared Max Context · higher is betterTest setup not documented as identicalGPT-5.4: 66.8%Kimi K3: 72.6%Vals AIAs of 08/12/2026
LegalBenchIssue Tasks · higher is betterTest setup not documented as identicalGPT-5.4: 90.5%Kimi K3: 90.9%Vals AIAs of 10/13/2026
LiveCodeBenchMedium · higher is betterTest setup not documented as identicalGPT-5.4: 85.6%Kimi K3: 88.5%Vals AIAs of 09/01/2026
LMArena Code Arena WebDevReact development · higher is betterTest setup not documented as identicalGPT-5.4: 1,388Kimi K3: 1,673LMArenaAs of 11/15/2026
LMArena Text, Style ControlCoding · higher is betterTest setup not documented as identicalGPT-5.4: 1,521Kimi K3: 1,538LMArenaAs of 11/11/2026
MMLU ProChemistry · higher is betterTest setup not documented as identicalGPT-5.4: 91.1%Kimi K3: 91.8%Vals AIAs of 09/01/2026
ProgramBenchRaw Pass Rate · higher is betterTest setup not documented as identicalGPT-5.4: 68.6%Kimi K3: 62.8%Vals AIAs of 11/05/2026
SWE-bench Verified1-4 hours · higher is betterTest setup not documented as identicalGPT-5.4: 50.0%Kimi K3: 92.9%Vals AIAs of 09/01/2026
LegalBenchRhetoric Tasks · higher is betterTest setup not documented as identicalGPT-5.4: 82.8%Kimi K3: 81.9%Vals AIAs of 10/13/2026
LMArena Code Arena WebDevWeb development · higher is betterTest setup not documented as identicalGPT-5.4: 1,400Kimi K3: 1,658LMArenaAs of 11/15/2026
LMArena Text, Style ControlCreative writing · higher is betterTest setup not documented as identicalGPT-5.4: 1,440Kimi K3: 1,459LMArenaAs of 11/11/2026
MMLU ProComputer Science · higher is betterTest setup not documented as identicalGPT-5.4: 89.8%Kimi K3: 91.2%Vals AIAs of 09/01/2026
SWE-bench Verified15 min - 1 hour · higher is betterTest setup not documented as identicalGPT-5.4: 76.2%Kimi K3: 94.6%Vals AIAs of 09/01/2026
LegalBenchRule Tasks · higher is betterTest setup not documented as identicalGPT-5.4: 84%Kimi K3: 84%Vals AIAs of 10/13/2026
LMArena Text, Style ControlEnglish · higher is betterTest setup not documented as identicalGPT-5.4: 1,476Kimi K3: 1,493LMArenaAs of 11/11/2026
MMLU ProEconomics · higher is betterTest setup not documented as identicalGPT-5.4: 89.7%Kimi K3: 91.6%Vals AIAs of 09/01/2026
LMArena Text, Style ControlEntertainment, sports and media · higher is betterTest setup not documented as identicalGPT-5.4: 1,441Kimi K3: 1,459LMArenaAs of 11/11/2026
MMLU ProEngineering · higher is betterTest setup not documented as identicalGPT-5.4: 85.2%Kimi K3: 86.9%Vals AIAs of 09/01/2026
LMArena Text, Style ControlExcluding ties · higher is betterTest setup not documented as identicalGPT-5.4: 1,484Kimi K3: 1,501LMArenaAs of 11/11/2026
MMLU ProHealth · higher is betterTest setup not documented as identicalGPT-5.4: 82.4%Kimi K3: 82.0%Vals AIAs of 09/01/2026
LMArena Text, Style ControlExpert prompts · higher is betterTest setup not documented as identicalGPT-5.4: 1,514Kimi K3: 1,528LMArenaAs of 11/11/2026
MMLU ProHistory · higher is betterTest setup not documented as identicalGPT-5.4: 79.3%Kimi K3: 79.3%Vals AIAs of 09/01/2026

More comparisons

Compare GPT-5.4 and Kimi K3 with the models people compare them with most often.

Compare another matchup

The most popular comparisons

  1. 1Claude Fable 5.2 vs. Kimi K3
  2. 2Claude Opus 5.5 vs. Kimi K3
  3. 3Claude Sonnet 5.5 vs. Kimi K3
  4. 4Claude Opus 5 vs. Kimi K3
  5. 5GPT-6 Astra vs. Kimi K3
  6. 6GPT-6 Luna vs. GPT-5.4
  7. 7Grok 4.7 vs. Kimi K3
  8. 8DeepSeek V4.1 Flash 1017 vs. Kimi K3
  9. 9DeepSeek-V4-Pro vs. Kimi K3
Show all 36 comparisons
  1. 10GPT-5.4 vs. GLM-5.3-Flash
  2. 11DeepSeek-V4.1-Flash vs. Kimi K3
  3. 12DeepSeek-V4-Flash vs. Kimi K3
  4. 13Kimi K3 vs. GLM-5.3-Flash
  5. 14Kimi K3 vs. GLM-5.3
  6. 15GPT-6 Sol vs. GPT-5.4
  7. 16GPT-5.6 Luna vs. Kimi K3
  8. 17Claude Sonnet 5 vs. Kimi K3
  9. 18Kimi K3 vs. Qwen 3.8 27B
  10. 19GPT-6 Sol vs. Kimi K3
  11. 20Gemini 3.1 Pro Preview vs. Kimi K3
  12. 21Muse Spark 1.3 vs. Kimi K3
  13. 22GPT-5.6 Luna vs. GPT-5.4
  14. 23GPT-5.4 vs. GLM-5.3
  15. 24Claude Sonnet 5.5 vs. GPT-5.4
  16. 25GPT-6 Astra vs. GPT-5.4
  17. 26Claude Opus 5.5 vs. GPT-5.4
  18. 27Claude Opus 4.8 vs. Kimi K3
  19. 28GPT-5.6 Terra vs. GPT-5.4
  20. 29Kimi K3 vs. Mistral Large 4
  21. 30GPT-5.6 Sol vs. GPT-5.4
  22. 31Grok 4.5 vs. Kimi K3
  23. 32Claude Fable 5 vs. GPT-5.4
  24. 33GPT-6.1 Sol vs. GPT-5.4
  25. 34Gemini 4 Argon vs. Kimi K3
  26. 35Muse Wonder vs. Kimi K3
  27. 36GPT-6.1 Terra vs. GPT-5.4