BenchLeader
Menu
Best LLMsChoose a modelCoding plansLeaderboardsModelsBenchmarksProvidersReleases12

Claude Opus 5.5 vs Kimi K3

Full profiles: Claude Opus 5.5 (Anthropic) and Kimi K3 (Moonshot AI): every benchmark, price and rank.

Updated · Data from 14 sources

As of 9 Nov 2026, Claude Opus 5.5 is ahead of Kimi K3 on quality (68.9 vs 62.0 on the BenchLeader Index), but Kimi K3 is 1.3× cheaper at $6.00 vs $8.00 per 1M tokens blended. Choose Claude Opus 5.5 for coding (80 vs 65), knowledge (74 vs 60), and reasoning (72 vs 60), where it also streams ~2× faster; choose Kimi K3 for fast first answers (65.8 s vs 380.6 s) and long context (69 vs 67).

Key facts (as of 9 Nov 2026)

  • Claude Opus 5.5 scores 68.9 on the BenchLeader Index and Kimi K3 scores 62.0 (BenchLeader, 9 Nov 2026).
  • Claude Opus 5.5 costs $8.00 per 1M tokens blended (list price, 9 Nov 2026).
  • Kimi K3 costs $6.00 per 1M tokens blended (list price, 9 Nov 2026).
  • Claude Opus 5.5 streams ~2× faster than Kimi K3 (Artificial Analysis, 7-day median, 9 Nov 2026).
  • Of the 43 benchmarks both models have, Claude Opus 5.5 leads 36 (BenchLeader, 9 Nov 2026).
  • On Arena WebDev, Claude Opus 5.5 is rated 1818 and Kimi K3 1657 (LMArena, as published by 9 Nov 2026).
Claude Opus 5.5 benchmarks max Anthropic Kimi K3 benchmarks max Moonshot AIAdd a third model

Which to pick for…

The pick's numbers come first. Category scores on the BenchLeader scale (50 is the average model, 15 points one standard deviation). High confidence means a lead of half a standard deviation or more; under one point is called either.

Use casePickWhyConfidence
CodingClaude Opus 5.580 vs 65 category score; Arena WebDev 1818 vs 1657High
Agents & tool useClaude Opus 5.563 vs 59 category score; Arena Agent 20 vs 0.6Medium
MathsClaude Opus 5.574 vs 63 category score; FrontierMath Tiers 1–3 91.2% vs 72.2%High
ReasoningClaude Opus 5.572 vs 60 category score; HLE-Diamond (AA) 67.9% vs 26.9%High
KnowledgeClaude Opus 5.574 vs 60 category score; SimpleQA Verified 72.2% vs 50.6%High
Long documentsKimi K369 vs 67 category score; AA-LCR v1.1 88.7% vs 84.7%Medium
Images and other mediaClaude Opus 5.570 vs 63 category score; MMMU-Pro 87.7% vs 80.5%Medium
Chat quality (human preference)–Only Kimi K3 has an Arena rating (68 category score)Not comparable
Lowest costKimi K3$6.00 vs $8.00 per 1M blended; agentic coding session $0.240 vs $0.320–
Fastest first answerKimi K365.8 s (Kimi K3 (max)) vs 380.6 s (Claude Opus 5.5 (max)) to the first answer token–
High volume on a budgetNeither flagshipA smaller model is usually the better fit: Cheapest good LLMs, Fastest LLMs–

Key differences

  • Claude Opus 5.5 (max) leads on quality: 68.9 vs 62.0.
  • Claude Opus 5.5 (max) is stronger in agents & tools, coding, knowledge, maths, multimodal, reasoning.
  • Kimi K3 (max) is stronger in long context, human preference.
  • Kimi K3 (max) is 1.3× cheaper ($6.00 vs $8.00 per 1M blended).
  • Claude Opus 5.5 (max) streams 2.7× faster (115 vs 42 tokens per second).

At a glance

SpecClaude Opus 5.5Kimi K3
DeveloperAnthropicMoonshot AI
Released22 Sep 202616 Jul 2026
Context window (tokens)1,000,0001,048,576
Max output (tokens)128,000 (up to 300K on the Message Batches API (beta))–
Input typestext, image–
Reasoning effort settingslow, medium, high, xhigh, max–
Compared atClaude Opus 5.5 (max)Kimi K3 (max)
Open weightsNoYes
Price in / out / cached input per 1M$4.00 / $20.00 / $0.200$3.00 / $15.00 / $0.300
API model IDclaude-opus-5-5–
Where to use itClaude API, Amazon Bedrock, Google Cloud, Microsoft Foundry–

Limits, availability and announced prices from the vendors' own pages: platform.claude.com (checked 20 Oct 2026). Everything else is from the day's data.

Category by category

Category scores on the BenchLeader scale: 50 is the average evaluated model, 15 points is one standard deviation. A gap under one point is called level.

CategoryClaude Opus 5.5 (max)Kimi K3 (max)Leader
Reasoning72.359.5Claude Opus 5.5
Coding80.464.5Claude Opus 5.5
Agents & tools62.758.7Claude Opus 5.5
Maths74.063.2Claude Opus 5.5
Knowledge73.760.0Claude Opus 5.5
Human preference–68.0not comparable
Multimodal70.363.0Claude Opus 5.5
Long context67.169.2Kimi K3

On one chart

Each domain on the index's own scale, where 50 is an average model. Click a domain to see the benchmarks behind it. Pick up to 5 models from the list.

2550 avg75ReasoningCodingAgents & toolsMathsKnowledgeLong contextMultimodalClaude Opus 5.5, Reasoning: 72.3Claude Opus 5.5, Coding: 80.4Claude Opus 5.5, Agents & tools: 62.7Claude Opus 5.5, Maths: 74.0Claude Opus 5.5, Knowledge: 73.7Claude Opus 5.5, Long context: 67.1Claude Opus 5.5, Multimodal: 70.3Kimi K3, Reasoning: 59.5Kimi K3, Coding: 64.5Kimi K3, Agents & tools: 58.7Kimi K3, Maths: 63.2Kimi K3, Knowledge: 60.0Kimi K3, Long context: 69.2Kimi K3, Multimodal: 63.0

Not drawn, measured for fewer than two of these models: Instruction following, Human preference.

ModelIndex

Where sources disagree

Across the 43 benchmarks both have, Claude Opus 5.5 leads 36, Kimi K3 leads 7. Every overall index that rates both puts Claude Opus 5.5 ahead.

Overall indexClaude Opus 5.5 (max)Kimi K3 (max)Leader
BenchLeader Index68.962.0Claude Opus 5.5
AA Intelligence Index v5.059.117.4†Claude Opus 5.5
Vals Index70.1†53.0Claude Opus 5.5
Arena Text1506†1487Claude Opus 5.5

Each publisher runs its own tasks, harnesses and effort settings, so their overall verdicts can differ; BenchLeader reconciles them on one scale. † measured at another effort setting of the same model (the publisher did not test the one compared here).

Price, speed and latency

List prices blended at three input tokens per output token. Per-task prices use the token counts on the methodology page. Cost, tokens and time per benchmark task are measured: Artificial Analysis’ averages over its whole Intelligence Index run.

MeasureClaude Opus 5.5 (max)Kimi K3 (max)
Blended price per 1M tokens$8.00$6.00
Input / output per 1M$4.00 / $20.00$3.00 / $15.00
Cached input per 1M$0.200$0.300
Price per task: chat reply$0.0076$0.0057
Price per task: summarize a 30-page report$0.060$0.045
Price per task: code edit$0.054$0.041
Price per task: agentic coding session$0.320$0.240
Price per task: structured extraction$0.012$0.0090
Agentic coding session, input mostly cached$0.149$0.118
Cost per benchmark task (AA Index)$10.04–
Output tokens per task (AA Index)173k–
Of which reasoning120k–
Time per task (AA Index)25 min–
Output speed115 tok/s42 tok/s
Time to first answer380.6 s65.8 s
Time to a complete answer771.1 s140.2 s

Each model is compared at its best-ranked reasoning-effort setting. Claude Opus 5.5 (max) scores 68.9 and takes 381 s to its first answer token; Claude Opus 5.5 (xhigh) scores 65.5 in 84 s, Claude Opus 5.5 (high) scores 65.4 in 28 s. Kimi K3 (max) scores 62.0 and takes 66 s to its first answer token; Kimi K3 scores 59.5 in 1 s, Kimi K3 (low) scores 53.7 in 12 s. Higher effort means more thinking before the answer: better scores, longer waits and more tokens billed. How reasoning effort changes scores and cost.

Every benchmark

160 benchmarks, grouped by category; each name links to the board and its leaders. † measured at another effort setting of the same model, shown for reference and not counted as a win.

MetricPublisherClaude Opus 5.5 (max)Kimi K3 (max)
BenchLeader IndexBenchLeader68.962.0
Agents & tools scoreBenchLeader62.758.7
Coding scoreBenchLeader80.464.5
Knowledge scoreBenchLeader73.760.0
Long context scoreBenchLeader67.169.2
Maths scoreBenchLeader74.063.2
Multimodal scoreBenchLeader70.363.0
Reasoning scoreBenchLeader72.359.5
Human preference scoreBenchLeader–68.0
Blended price $/MBenchLeader$8.00$6.00
Output speedBenchLeader115 tok/s42 tok/s
Time to first answerBenchLeader380.6 s65.8 s
Context windowBenchLeader1,000,000 tokens1,048,576 tokens
Reasoning
GPQA DiamondEpoch AI Benchmarking Hub90.6%93.1%
SimpleBenchSimpleBench–60.7%
Arena Hard PromptsArena (LMArena)1530†1515
LiveBench 2.0 ReasoningLiveBench92.2%†90.7%†
LiveBench 1.0 ReasoningLiveBench92.2%90.7%†
GPQA Diamond (AA)Artificial Analysis–93.5%
Humanity's Last Exam (AA)Artificial Analysis61.4%46.9%
HLE-Diamond (AA)Artificial Analysis67.9%26.9%
GPQA Diamond (Vals)Vals AI–92.9%†
ARC-AGI-1ARC Prize97.5%94.5%
ARC-AGI-2ARC Prize91.7%60.4%
CritPtArtificial Analysis31.7%23.4%
CritPt v2 hard subset (AA)Artificial Analysis68.0%40.5%
MysteryMechanismVals AI49.5%†–
Chess PuzzlesEpoch AI Benchmarking Hub67.0%36.0%
Mystery Game PuzzlesEpoch AI Benchmarking Hub71.0%26.0%
LMCAEpoch AI Benchmarking Hub68.2%52.7%
DTBenchEpoch AI Benchmarking Hub98.9%91.2%
ForecastBenchForecastBench–64.7%
Coding
SciCodeSciCode66.9%59.5%
WeirdMLWeirdML–82.6%
FrontierCodeCognition54.6%†44.2%†
Arena CodingArena (LMArena)1546†1537
Arena WebDevArena (LMArena)18181657
LiveBench 2.0 CodingLiveBench87.1%†53.1%†
Show all 160 benchmarks
MetricPublisherClaude Opus 5.5 (max)Kimi K3 (max)
Coding
LiveBench 1.0 CodingLiveBench89.3%81.5%†
LiveCodeBenchVals AI–87.2%
IOIVals AI95.1%†48.9%
SWE-bench (Vals)Vals AI–93.4%†
Code MigrationVals AI66.7%†16.1%
Code Migration v1.1Vals AI81.0%†28.6%
ProgramBenchVals AI18.5%†2.0%†
SWE-bench MultimodalVals AI70.2%†44.2%
Vibe Code Bench 1-100Vals AI30.4%†18.2%
Vibe Code Bench v1.1Vals AI90.3%†85.0%
MirrorCodeEpoch AI Benchmarking Hub78.9%†–
Surface Evolver BenchEpoch AI Benchmarking Hub–93.0%
DeepSWE v1.1Epoch AI Benchmarking Hub–68.5%
CursorBenchCursor57.8%–
ALE-BenchEpoch AI Benchmarking Hub2147.4†1524.5
FrontierSWEEpoch AI Benchmarking Hub62.3%25.9%
Terminal-Bench 4.0 (AA)Artificial Analysis59.6%12.6%
Terminal-Bench 2.1 (AA)Artificial Analysis–85.0%
Startrise FrontendStartrise71.1%†79.8%†
GameBench 2 OverallDuelLab GameBench 287.6†47.2†
GameBench 2 by reasoning settingDuelLab GameBench 294.8†43.8
Design Arena WebsiteDesign Arena1325†1332†
FrontierCode 1.1 (AA)Artificial Analysis52.2%43.6%
FrontierCode 2.0 (AA)Artificial Analysis46.4%10.9%†
Agents & tools
Terminal-BenchTerminal-Bench65.8%–
Remote Labor IndexScale AI / CAIS21.3%†–
APEX-AgentsMercor73.5%47.8%†
Arena AgentArena (LMArena)200.6
LiveBench Long AgenticLiveBench72.8%†36.0%†
LiveBench BrowserLiveBench88.4%†75.2%†
LiveBench Agentic CodingLiveBench71.7%62.2%†
Terminal-Bench 2.1 (Vals)Vals AI87.6%†80.9%†
MCP AtlasScale AI SEAL85.4%†82.3%
HiL-BenchScale AI SEAL59.0%†42.7%†
τ³-bench Bankingτ²-bench–37.1%
GDPval-AA v2.1, normalised scoreArtificial Analysis68.3%–
GDPval-AA v2.1Artificial Analysis1866–
τ³-Banking (AA)Artificial Analysis–46.0%
ITBench SRE (AA)Artificial Analysis38.2%47.7%
Analyst Agent (AA)Artificial Analysis56.3%38.8%
APEX-Agents (AA)Artificial Analysis–41.3%
CUA-benchVals AI14.0%–
CyberBenchVals AI55.4%†75.2%
Finance Agent v2Vals AI58.6%†53.1%
Finance Agent v2.1Vals AI67.8%†60.0%
Harvey's Legal Agent Benchmark v1.1Vals AI9.2%†12.9%
Harvey's Legal Agent Benchmark v1Vals AI3.8%†12.9%
Legal Research BenchVals AI50.5%†46.1%
Robo-benchVals AI14.6%–
SREBenchVals AI33.6%†–
Tax Agent BenchVals AI44.5%†33.4%
Terminal-Bench 4.0 (Vals)Vals AI65.2%†17.2%
Terminal-Bench ScienceVals AI47.1%†1.4%
Time Horizon Index: KSPVals AI91.3%†10.5%
Arena SearchArena (LMArena)1214†–
PostTrainBenchEpoch AI Benchmarking Hub–32.0%†
GBAEvalEpoch AI Benchmarking Hub–48.4%†
Vending-Bench 2Andon Labs9235.3†5165.0†
AutomationBenchArtificial Analysis69.5%58.3%
GDP.pdf (AA)Artificial Analysis26.2%22.0%
Harvey LAB-AA v1.1Harvey4.2%5.3%
EnterpriseOps-GymArtificial Analysis–45.3%
AA-Briefcase v1.1Artificial Analysis18001497
GDPxGDPx by SemiAnalysis12.0%3.7%
Design Arena AgenticDesign Arena1334†1257†
Terminal-Bench-Science 0.1 (AA)Artificial Analysis59.0%8.1%
CWE-Bench (AA)Artificial Analysis58.3%53.3%
DeepSec-Bench (AA)Artificial Analysis26.9%12.4%
CyberGym end-to-end (AA)Artificial Analysis0.8%58.0%
FACTS SearchGoogle88.4%†90.3%†
Maths
FrontierMath Tiers 1–3Epoch AI Benchmarking Hub91.2%72.2%
FrontierMath Tier 4Epoch AI Benchmarking Hub95.0%39.0%
OTIS Mock AIMEEpoch AI Benchmarking Hub100.0%97.2%
ProofBenchVals AI100.0%87.0%†
LiveBench MathematicsLiveBench97.1%84.4%†
AIME 2026MathArena–96.7%†
HMMT February 2026MathArena–97.0%†
MathArena ApexMathArena–65.6%†
Arena MathsArena (LMArena)1518†1502
Knowledge
SimpleQA VerifiedEpoch AI Benchmarking Hub72.2%50.6%
LiveBench 2.0 Data AnalysisLiveBench80.3%†78.7%†
LiveBench ResearchLiveBench88.9%†79.8%†
LiveBench 1.0 Data AnalysisLiveBench80.3%78.7%†
AA-Omniscience v1.1Artificial Analysis48.1-1.7†
MMLU-ProVals AI–88.0%
LegalBenchVals AI–86.2%
CorpFinVals AI–71.6%†
TaxEvalVals AI–75.7%†
PRBench LegalScale AI SEAL47.6%†–
BioMysteryBenchVals AI79.3%†72.6%
ClaimBenchVals AI100.0%†–
Excel Modeling BenchmarkVals AI75.9%†66.7%
Excel Modeling Benchmark v1.1Vals AI75.8%†71.3%
MedCodeVals AI49.8%†49.4%
MedScribeVals AI91.4%†88.0%
Public Benefits Bench v1.1Vals AI70.6%†68.2%
EBR-benchEpoch AI Benchmarking Hub71.4%–
GDP.pdfEpoch AI Benchmarking Hub30.6%19.0%
AA-Omniscience v1.1: accuracyArtificial Analysis65.5%42.4%†
AA-Omniscience v1.1: non-hallucinationArtificial Analysis49.7%23.4%†
HealthBench Professional (AA)OpenAI59.9%53.2%
Instruction following
LiveBench LanguageLiveBench86.3%85.5%†
LiveBench Instruction FollowingLiveBench65.7%71.4%†
Arena Instruction FollowingArena (LMArena)1519†1489
Human preference
Arena TextArena (LMArena)1506†1487
EQ-Bench 4EQ-Bench1340†1319†
Arena Creative WritingArena (LMArena)1517†1459
Arena Multi-turnArena (LMArena)1500†1501
Arena Longer QueriesArena (LMArena)1520†1500
Multimodal
Arena VisionArena (LMArena)1324†–
LiveBench DesignLiveBench83.3%†72.5%†
MMMU-ProArtificial Analysis87.7%80.5%
MMMU-Pro (Vals)Vals AI–88.2%†
MortgageTaxVals AI–66.3%†
SAGEVals AI91.3%†86.8%
Blueprint-Bench 2Epoch AI Benchmarking Hub51.2%†29.5%†
Code2Video-AAArtificial Analysis15361041
AA-3DArtificial Analysis1202984
Long context
AA-LCR v1.1Artificial Analysis84.7%88.7%
MLCRArtificial Analysis27.8%21.7%
Composite
Epoch Capabilities IndexEpoch AI Benchmarking Hub168.4†158.5†
LiveBench 2.0LiveBench83.4%†65.0%†
LiveBench 1.0LiveBench83.2%79.2%†
AA Intelligence Index v5.0Artificial Analysis59.117.4†
AA Intelligence Index v4.3.2Artificial Analysis57.643.6
AA Progress Index v1Artificial Analysis350319.9
Vals IndexVals AI70.1†53.0
Vals Finance IndexVals AI65.3%†58.0%
Vals RSI IndexVals AI37.3%20.9%
Vals Science IndexVals AI58.6%†–
Vals Multimodal IndexVals AI–73.4%†
Vals Router IndexVals AI64.2%†53.0%†
AA Cyber IndexArtificial Analysis28.741.3
Safety & honesty
FORTRESSScale AI SEAL5.8%26.6%
Metadata
AA Openness IndexArtificial Analysis–38.9

Data as of 2026-11-09. Best configuration of each model; every score links to its source on the model pages.

DuelLab GameBench 2, release gb2-20261103-eaea31df94e6 (https://benchmarks.duellab.org/releases/gb2-20261103-eaea31df94e6/), CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Scores mapped to BenchLeader model names; not endorsed by DuelLab.

What changed

  • Claude Opus 5.5 (max): 2 score revisions.
  • Claude Opus 5.5 (max): 1 score revision.
  • Claude Opus 5.5 (max): 1 score revision.
  • Claude Opus 5.5 (max): 2 score revisions.
  • Kimi K3 (max): 2 score revisions.
  • Claude Opus 5.5 (max): 2 score revisions.
  • Claude Opus 5.5 (max): 1 score revision.
  • Kimi K3 (max): 1 new result (MirrorCode).
  • Kimi K3 (max): 1 score revision.
  • Kimi K3 (max): BenchLeader Index 63.1 → 62.1.

Rankings: Best LLM overall · Best LLM for coding · Best LLM for agents · Best cheap LLMs

Claude Opus 5.5 vs Kimi K3: questions

Is Claude Opus 5.5 better than Kimi K3?
Claude Opus 5.5 (max) leads on quality: 68.9 vs 62.0. The BenchLeader Index combines every independent quality benchmark; Claude Opus 5.5 (max) is ahead overall as of 9 Nov 2026, but check the category scores for your use.
Claude Opus 5.5 vs Kimi K3: which is cheaper, and will that change?
Kimi K3 is cheaper today: $6.00 against $8.00 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
Is Claude Opus 5.5 or Kimi K3 better for coding?
Claude Opus 5.5 scores higher in coding (80 vs 65 on the category index, where 50 is average). On Arena WebDev, the top shared coding benchmark, it is 1818 for Claude Opus 5.5 against 1657 for Kimi K3.
Which is faster, Claude Opus 5.5 or Kimi K3?
Claude Opus 5.5 streams ~2× faster (115 vs 42 output tokens per second, Artificial Analysis 7-day median).
Is Claude Opus 5.5 or Kimi K3 better for agents and tool use?
Claude Opus 5.5 scores higher in agentic tasks (63 vs 59 on the category index, where 50 is average). On Arena Agent, the top shared agents & tools benchmark, it is 20 for Claude Opus 5.5 against 0.6 for Kimi K3.
More questions (4)
Which is better for long documents?
Kimi K3 scores higher on long-context benchmarks (69 vs 67). Both take about 1M tokens: Claude Opus 5.5 1,000,000, Kimi K3 1,048,576.
Which has lower latency?
Kimi K3 answers first: 65.8 s against 380.6 s to the first answer token, as Artificial Analysis measures it. Reasoning effort drives most of this: a higher setting thinks longer before it answers.
Do independent benchmarks agree on Claude Opus 5.5 vs Kimi K3?
Across the 43 benchmarks both have, Claude Opus 5.5 leads 36, Kimi K3 leads 7. Every overall index that rates both puts Claude Opus 5.5 ahead.
When was Claude Opus 5.5 released?
Claude Opus 5.5 was released on 22 Sep 2026.