Enter opens the highlighted result. Esc closes.

GPT-5.2 vs Kimi K3

Kimi K3 leads GPT-5.2 on coding, 62.3 to 55.9 on the Vibe Coding Index, at about the same output price ($15.00 vs $14.00 per million). Updated October 22, 2026; 77 benchmarks have a result for both.

Replace GPT-5.2 Replace Kimi K3
GPT-5.2OpenAI
Coding
55.9
Overall
60.7
Output $/M
$14.00
5
Kimi K3Moonshot AI
Coding
62.3
Overall
66.5
Output $/M
$15.00
72
vs boards won of 77 shared

GPT-5.2 or Kimi K3: which to pick

Pick Kimi K3 for coding: it scores 62.3 on the Vibe Coding Index against 55.9 for GPT-5.2.

Pick GPT-5.2 for output tokens 1.1x cheaper ($14.00 vs $15.00 per million) and work measured by Chess puzzles or TaxEval (Vals AI), where it leads.

Model AGPT-5.2OpenAI55.9/1003 independent benchmarks$1.91 per solved taskModel BKimi K3Moonshot AI62.3/1006 independent benchmarks$0.81 per solved task
Shared results
77
GPT-5.2 only
37
Kimi K3 only
40
Like-for-like categories
8 of 9

Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.

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-5.2 55.9 from 3 benchmarks, Kimi K3 62.3 from 6 benchmarks.
Widest category
Web design, 22.6 points apart on average over 7 shared benchmarks.
Widest single gap
ProofBench: GPT-5.2 15.0% against Kimi K3 87.0%, 72.0 points apart on a 0 to 100 scale.
Output price
GPT-5.2 $14.00 against Kimi K3 $15.00 per million output tokens, 1.1x apart.
Context window
GPT-5.2 takes 400K tokens, Kimi K3 1.05M.

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-5.2BarsKimi K3Ties
Coding09
Intelligence02
Agentic work06
Reasoning112
Math06
Knowledge38
Web design07
General122
All shared boards572
CodingLike-for-likeSix of these boards feed the Vibe Coding Index09
Coding benchmarks for GPT-5.2 and Kimi K3
BenchmarkGPT-5.2Kimi K3Leader
SWE-bench Verified72.8%Not measuredOnly GPT-5.2 measured
SWE-bench Multilingual66.7%Not measuredOnly GPT-5.2 measured
LiveBench codingNot measured59.3%Only Kimi K3 measured
LMArena WebDev14171655Kimi K3
AA Coding IndexNot measured76.2Only Kimi K3 measured
LiveCodeBench (Vals AI)85.4%87.2%Kimi K3
SWE-bench Verified (Vals AI)75.8%93.4%Kimi K3
Terminal-Bench 4 (Vals AI)Not measured17.2%Only Kimi K3 measured
Vibe Code Bench (Vals AI)53.5%85.0%Kimi K3
Code migration (Vals AI)Not measured16.1%Only Kimi K3 measured
ProgramBench (Vals AI)Not measured2.0%Only Kimi K3 measured
IOI (Vals AI)Not measured48.9%Only Kimi K3 measured
Terminal-Bench 2.1 (Vals AI)Not measured80.9%Only Kimi K3 measured
SWE-bench Multimodal (Vals AI)Not measured44.2%Only Kimi K3 measured
Vibe Code Bench 1-100 (Vals AI)Not measured18.2%Only Kimi K3 measured
CyberBench (Vals AI)Not measured75.2%Only Kimi K3 measured
LiveCodeBench (Artificial Analysis run)89.4%Not measuredOnly GPT-5.2 measured
SciCodeNot measured59.5%Only Kimi K3 measured
Terminal-Bench Hard (Artificial Analysis run)47.0%Not measuredOnly GPT-5.2 measured
Terminal-Bench64.9%Not measuredOnly GPT-5.2 measured
FrontierCodeNot measured44.2%Only Kimi K3 measured
FrontierSWENot measured25.9%Only Kimi K3 measured
GSO-Bench27.5%Not measuredOnly GPT-5.2 measured
WeirdML72.2%82.6%Kimi K3
ALE-Bench1293.51524.5Kimi K3
LMArena Text: coding14421508Kimi K3
LMArena Text: software and IT14341499Kimi K3
LMArena WebDev: ReactNot measured1670Only Kimi K3 measured
LMArena WebDev: HTML14271644Kimi K3
LMArena WebDev: from an imageNot measured1579Only Kimi K3 measured
Neither model has a result on the other 9 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index02
Intelligence benchmarks for GPT-5.2 and Kimi K3
BenchmarkGPT-5.2Kimi K3Leader
AA Intelligence Index30.443.6Kimi K3
Epoch Capabilities Index153.7157.9Kimi K3
Agentic workLike-for-likePublic boards, shown as published, kept out of the index06
Agentic work benchmarks for GPT-5.2 and Kimi K3
BenchmarkGPT-5.2Kimi K3Leader
AA Agentic IndexNot measured50.0Only Kimi K3 measured
Design Arena: agents building web apps10861295Kimi K3
Design Arena: agents building full-stack apps10431315Kimi K3
Design Arena: agents building mobile apps10931269Kimi K3
Design Arena: agents building Android apps10591253Kimi K3
Design Arena: agents building Godot games11421223Kimi K3
Design Arena: agents building gamesNot measured1252Only Kimi K3 measured
Design Arena: HTML slidesNot measured1206Only Kimi K3 measured
Design Arena: PowerPoint slidesNot measured1247Only Kimi K3 measured
Design Arena: spreadsheetsNot measured1070Only Kimi K3 measured
LiveBench researchNot measured79.0%Only Kimi K3 measured
LiveBench browserNot measured16.0%Only Kimi K3 measured
LiveBench long agenticNot measured42.8%Only Kimi K3 measured
Finance Agent (Vals AI)Not measured53.1%Only Kimi K3 measured
Tax Agent Bench (Vals AI)Not measured33.4%Only Kimi K3 measured
Legal Research Bench (Vals AI)Not measured46.1%Only Kimi K3 measured
Harvey's Legal Agent Benchmark (Vals AI)Not measured12.9%Only Kimi K3 measured
Excel Modeling Benchmark (Vals AI)Not measured66.7%Only Kimi K3 measured
BioMysteryBench (Vals AI)Not measured72.6%Only Kimi K3 measured
Terminal-Bench-Science (Vals AI)Not measured1.4%Only Kimi K3 measured
Time Horizon Index (Vals AI)Not measured10.5%Only Kimi K3 measured
τ²-Bench Telecom84.8%Not measuredOnly GPT-5.2 measured
APEX-AgentsNot measured47.8%Only Kimi K3 measured
METR time horizons75.3%Not measuredOnly GPT-5.2 measured
Remote Labor Index2.5%Not measuredOnly GPT-5.2 measured
GDPval49.7%Not measuredOnly GPT-5.2 measured
DeepResearch Bench41.1%Not measuredOnly GPT-5.2 measured
PostTrainBenchNot measured32.0%Only Kimi K3 measured
Vending-Bench 2$3,591$5,165Kimi K3
Berkeley Function Calling Leaderboard (BFCL)55.9%Not measuredOnly GPT-5.2 measured
Neither model has a result on the other 11 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index112
Reasoning benchmarks for GPT-5.2 and Kimi K3
BenchmarkGPT-5.2Kimi K3Leader
GPQA Diamond91.7%92.9%Kimi K3
LiveBench reasoningNot measured90.7%Only Kimi K3 measured
MMMU Pro (Vals AI)86.7%88.2%Kimi K3
GPQA Diamond (Artificial Analysis run)90.3%93.5%Kimi K3
Humanity's Last Exam (Artificial Analysis run)37.7%46.9%Kimi K3
GPQA Diamond (Epoch AI run)91.4%93.1%Kimi K3
ARC-AGI86.2%94.5%Kimi K3
ARC-AGI-252.9%60.4%Kimi K3
SimpleBench45.8%60.7%Kimi K3
Chess puzzles48.0%37.0%GPT-5.2
Mystery game puzzles23.0%26.0%Kimi K3
EBR-bench23.0%Not measuredOnly GPT-5.2 measured
VPCT84.0%Not measuredOnly GPT-5.2 measured
LMArena Text: hard prompts14281495Kimi K3
LMArena Text: expert14451520Kimi K3
LMArena Text: hard prompts in English14311496Kimi K3
LMArena Vision: diagrams1265Not measuredOnly GPT-5.2 measured
LMArena Vision: homework1276Not measuredOnly GPT-5.2 measured
Kagi LLM Benchmark73.3%Not measuredOnly GPT-5.2 measured
Neither model has a result on the other 3 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index06
Math benchmarks for GPT-5.2 and Kimi K3
BenchmarkGPT-5.2Kimi K3Leader
AIME96.9%Not measuredOnly GPT-5.2 measured
MGSM (Vals AI)94.0%Not measuredOnly GPT-5.2 measured
AA Math Index99.0Not measuredOnly GPT-5.2 measured
AIME 2025 (Artificial Analysis run)99.0%Not measuredOnly GPT-5.2 measured
OTIS Mock AIME96.1%97.2%Kimi K3
FrontierMath, tiers 1 to 367.4%72.2%Kimi K3
FrontierMath, tier 431.7%39.0%Kimi K3
ProofBench15.0%87.0%Kimi K3
LMArena Text: math14401494Kimi K3
LMArena Text: mathematical work14421505Kimi K3
Neither model has a result on the other 3 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index38
Knowledge benchmarks for GPT-5.2 and Kimi K3
BenchmarkGPT-5.2Kimi K3Leader
MMLU Pro86.2%88.0%Kimi K3
LegalBench (Vals AI)82.8%86.2%Kimi K3
CaseLaw (Vals AI)66.0%Not measuredOnly GPT-5.2 measured
CorpFin (Vals AI)65.9%71.6%Kimi K3
TaxEval (Vals AI)75.8%75.7%GPT-5.2
MortgageTax (Vals AI)67.1%66.3%GPT-5.2
MedQA (Vals AI)94.1%Not measuredOnly GPT-5.2 measured
MedCode (Vals AI)49.8%49.4%GPT-5.2
MMLU Pro (Artificial Analysis run)87.4%Not measuredOnly GPT-5.2 measured
SimpleQA Verified37.1%50.6%Kimi K3
LMArena Text: science14251508Kimi K3
LMArena Text: medicine and healthcare14251489Kimi K3
LMArena Text: legal and government14271503Kimi K3
LMArena Text: business and finance14141470Kimi K3
FACTS Benchmark Suite48.5%Not measuredOnly GPT-5.2 measured
Neither model has a result on the other 3 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index07
Web design benchmarks for GPT-5.2 and Kimi K3
BenchmarkGPT-5.2Kimi K3Leader
Design Arena: websites11951334Kimi K3
Design Arena: UI components11901341Kimi K3
Design Arena: game dev11951358Kimi K3
Design Arena: data viz12021351Kimi K3
Design Arena: 3D10791387Kimi K3
Design Arena: all code categories11701360Kimi K3
Design Arena: SVG11391302Kimi K3
Design Arena: ASCII art1152Not measuredOnly GPT-5.2 measured
LiveBench designNot measured77.2%Only Kimi K3 measured
GeneralLike-for-likePublic boards, shown as published, kept out of the index122
General benchmarks for GPT-5.2 and Kimi K3
BenchmarkGPT-5.2Kimi K3Leader
Vals IndexNot measured53.0%Only Kimi K3 measured
LiveBench data analysisNot measured78.7%Only Kimi K3 measured
Public Benefits Bench (Vals AI)Not measured68.2%Only Kimi K3 measured
SAGE (Vals AI)49.3%52.8%Kimi K3
MedScribe (Vals AI)84.4%88.0%Kimi K3
Vals Multimodal IndexNot measured73.4%Only Kimi K3 measured
IFBench75.4%Not measuredOnly GPT-5.2 measured
Long context reasoning (AA-LCR)82.7%88.7%Kimi K3
DTBench90.9%91.2%Kimi K3
LMCA43.9%52.7%Kimi K3
CL-bench18.2%Not measuredOnly GPT-5.2 measured
Surface Evolver BenchNot measured95.0%Only Kimi K3 measured
Furniture assembly38.3%34.2%GPT-5.2
LMArena Text14161476Kimi K3
LMArena Text: instruction following14091485Kimi K3
LMArena Text: creative writing13761457Kimi K3
LMArena Vision1246Not measuredOnly GPT-5.2 measured
LMArena Text: English14221481Kimi K3
LMArena Text: non-English14051466Kimi K3
LMArena Text: Chinese14471529Kimi K3
LMArena Text: German14441490Kimi K3
LMArena Text: Japanese14201494Kimi K3
LMArena Text: Korean13741460Kimi K3
LMArena Text: French14541492Kimi K3
LMArena Text: Spanish14131472Kimi K3
LMArena Text: Russian14151485Kimi K3
LMArena Text: Polish14261458Kimi K3
LMArena Text: multi-turn14191489Kimi K3
LMArena Text: longer queries14131494Kimi K3
LMArena Text: writing and language13921471Kimi K3
LMArena Text: entertainment and media13781453Kimi K3
LMArena Vision: English1243Not measuredOnly GPT-5.2 measured
LMArena Vision: Chinese1305Not measuredOnly GPT-5.2 measured
LMArena Vision: reading text1257Not measuredOnly GPT-5.2 measured
LMArena Vision: humor1227Not measuredOnly GPT-5.2 measured
LMArena Vision: captioning1267Not measuredOnly GPT-5.2 measured
LMArena Vision: recognizing things1232Not measuredOnly GPT-5.2 measured
LMArena Vision: creative writing1225Not measuredOnly GPT-5.2 measured
LMArena Document1405Not measuredOnly GPT-5.2 measured
Neither model has a result on the other 7 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for GPT-5.2 and Kimi K3
BenchmarkGPT-5.2Kimi K3Leader
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-5.2$0.0088Fits in one request
Kimi K3$0.011Fits in one request

GPT-5.2 costs less

Repository review

50K input + 3K output tokens

GPT-5.2$0.130Fits in one request
Kimi K3$0.195Fits in one request

GPT-5.2 costs less

Agent loop

200K input + 10K output tokens

GPT-5.2$0.490Fits in one request
Kimi K3$0.750Fits in one request

GPT-5.2 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-5.2 and Kimi K3
SpecificationOpenAI GPT-5.2MoonshotAI Kimi K3
MakerOpenAIMoonshot AI
WeightsNo public weights listedOpen weights
ReleasedDecember 10, 2025July 16, 2026
Context window400K1.05M
Input $/Mtok$1.75$3.00
Output $/Mtok$14.00$15.00
Coding index55.962.3
Independent measurements behind it36
Cost per solved task$1.91$0.81
SourceOpenRouter listingOpenRouter listing

Questions

Is GPT-5.2 better than Kimi K3 for coding?

Pick Kimi K3 for coding: it scores 62.3 on the Vibe Coding Index against 55.9 for GPT-5.2. Pick GPT-5.2 for output tokens 1.1x cheaper ($14.00 vs $15.00 per million) and work measured by Chess puzzles or TaxEval (Vals AI), where it leads.

Which is cheaper, GPT-5.2 or Kimi K3?

GPT-5.2 costs $1.75 per million input tokens and $14.00 per million output tokens; Kimi K3 costs $3.00 and $15.00. Per solved SWE-bench Verified task, GPT-5.2 costs $1.91 and Kimi K3 $0.81.

Which has the bigger context window, GPT-5.2 or Kimi K3?

Kimi K3, with 1.05M tokens against 400K.

How many benchmarks is the GPT-5.2 vs Kimi K3 comparison based on?

77 benchmarks have a result for both models as of October 22, 2026. GPT-5.2's coding index rests on 3 independent benchmarks and Kimi K3's on 6. Where only one model has a result, the page says so instead of naming a winner.

Compare next

Embed this comparison on your site

Free under CC BY 4.0, while the source link stays visible. Preview the frame · Terms and other embeds

<iframe src="https://llmbenchmark.ai/embed/compare/gpt-5-2-vs-kimi-k3" title="GPT-5.2 vs Kimi K3, LLMBenchmark.ai" width="100%" height="360" style="border:0;max-width:100%" loading="lazy"></iframe>
<p style="font:13px/1.4 sans-serif;margin:6px 0 0">Source: <a href="https://llmbenchmark.ai/compare/gpt-5-2-vs-kimi-k3">LLMBenchmark.ai, GPT-5.2 vs Kimi K3</a></p>