BenchLeader
Menu
Best LLMsChoose a modelCoding plansLeaderboardsModelsBenchmarksProvidersReleases8

Claude Opus 5 vs Kimi K2.5

Full profiles: Claude Opus 5 (Anthropic) and Kimi K2.5 (Moonshot AI): every benchmark, price and rank.

Updated · Data from 14 sources

As of 29 Oct 2026, Claude Opus 5 is ahead of Kimi K2.5 on quality (65.5 vs 58.8 on the BenchLeader Index), but Kimi K2.5 is 8.3× cheaper at $1.20 vs $10.00 per 1M tokens blended. Claude Opus 5 leads in every category we measure both in. Kimi K2.5 streams faster.

Key facts (as of 29 Oct 2026)

  • Claude Opus 5 scores 65.5 on the BenchLeader Index and Kimi K2.5 scores 58.8 (BenchLeader, 29 Oct 2026).
  • Claude Opus 5 costs $10.00 per 1M tokens blended (list price, 29 Oct 2026).
  • Kimi K2.5 costs $1.20 per 1M tokens blended (list price, 29 Oct 2026).
  • Kimi K2.5 streams faster than Claude Opus 5 (Artificial Analysis, 7-day median, 29 Oct 2026).
  • Of the 16 benchmarks both models have, Claude Opus 5 leads 16 (BenchLeader, 29 Oct 2026).
  • On Arena WebDev, Claude Opus 5 is rated 1690 and Kimi K2.5 1437 (LMArena, as published by 29 Oct 2026).
Claude Opus 5 benchmarks max Anthropic Kimi K2.5 benchmarks thinking Moonshot AIOpen in the comparison tool

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 570 vs 56 category score; Arena WebDev 1690 vs 1437High
Agents & tool useClaude Opus 564 vs 45 category score; τ³-Banking (AA) 42.1% vs 14.2%High
MathsClaude Opus 570 vs 56 category score; Arena Maths 1517 vs 1471High
ReasoningClaude Opus 569 vs 67 category score; Arena Hard Prompts 1515 vs 1471Medium
KnowledgeClaude Opus 565 vs 60 category scoreMedium
Long documentsEither65 vs 64 category score; AA-LCR v1.1 79.3% vs 78.0%Either
Images and other mediaClaude Opus 568 vs 59 category score; MMMU-Pro 84.7% vs 75.4%High
Chat quality (human preference)Claude Opus 568 vs 63 on Arena votesMedium
Lowest costKimi K2.5$1.20 vs $10.00 per 1M blended; agentic coding session $0.048 vs $0.400–
Fastest first answerEither43.9 s (Claude Opus 5 (max)) vs 36.0 s (Kimi K2.5 (thinking)) 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 (max) leads on quality: 65.5 vs 58.8.
  • Claude Opus 5 (max) is stronger in agents & tools, coding, human preference, knowledge, long context, maths, multimodal, reasoning.
  • Kimi K2.5 (thinking) is stronger in instruction following.
  • Kimi K2.5 (thinking) is 8.3× cheaper ($1.20 vs $10.00 per 1M blended).
  • Kimi K2.5 (thinking) streams 1.6× faster (87 vs 56 tokens per second).

At a glance

SpecClaude Opus 5Kimi K2.5
DeveloperAnthropicMoonshot AI
Released24 Jul 202627 Jan 2026
Context window (tokens)1,000,000262,144
Compared atClaude Opus 5 (max)Kimi K2.5 (thinking)
Open weightsNoYes
Price in / out / cached input per 1M$5.00 / $25.00 / $0.500$0.600 / $3.00 / $0.100

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 (max)Kimi K2.5 (thinking)Leader
Reasoning69.267.3Claude Opus 5
Coding70.055.7Claude Opus 5
Agents & tools64.345.1Claude Opus 5
Maths70.456.2Claude Opus 5
Knowledge65.060.0Claude Opus 5
Instruction following–69.0not comparable
Human preference68.463.4Claude Opus 5
Multimodal67.659.3Claude Opus 5
Long context64.563.8level

Where sources disagree

Across the 16 benchmarks both have, Claude Opus 5 leads 16, Kimi K2.5 leads 0. Every overall index that rates both puts Claude Opus 5 ahead.

Overall indexClaude Opus 5 (max)Kimi K2.5 (thinking)Leader
BenchLeader Index65.558.8Claude Opus 5
Arena Text14911450Claude Opus 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; suite cost is what Artificial Analysis paid to run its whole Intelligence Index.

MeasureClaude Opus 5 (max)Kimi K2.5 (thinking)
Blended price per 1M tokens$10.00$1.20
Input / output per 1M$5.00 / $25.00$0.600 / $3.00
Cached input per 1M$0.500$0.100
Price per task: chat reply$0.0095$0.0011
Price per task: summarize a 30-page report$0.075$0.0090
Price per task: code edit$0.068$0.0081
Price per task: agentic coding session$0.400$0.048
Price per task: structured extraction$0.015$0.0018
Output speed56 tok/s87 tok/s
Time to first answer43.9 s36.0 s
Time to a complete answer52.8 s41.7 s

Each model is compared at its best-ranked reasoning-effort setting. Claude Opus 5 (max) scores 65.5 and takes 44 s to its first answer token; Claude Opus 5 scores 65.2 in 3 s, Claude Opus 5 (high) scores 64.4 in 14 s. Kimi K2.5 (thinking) scores 58.8 and takes 36 s to its first answer token; Kimi K2.5 scores 54.3 in 1 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

146 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 (max)Kimi K2.5 (thinking)
BenchLeader IndexBenchLeader65.558.8
Agents & tools scoreBenchLeader64.345.1
Coding scoreBenchLeader70.055.7
Human preference scoreBenchLeader68.463.4
Knowledge scoreBenchLeader65.060.0
Long context scoreBenchLeader64.563.8
Maths scoreBenchLeader70.456.2
Multimodal scoreBenchLeader67.659.3
Reasoning scoreBenchLeader69.267.3
Instruction following scoreBenchLeader–69.0
Blended price $/MBenchLeader$10.00$1.20
Output speedBenchLeader56 tok/s87 tok/s
Time to first answerBenchLeader43.9 s36.0 s
Context windowBenchLeader1,000,000 tokens262,144 tokens
Reasoning
GPQA DiamondEpoch AI Benchmarking Hub93.9%87.6%†
Humanity's Last ExamScale AI / CAIS–24.4%†
SimpleBenchSimpleBench80.6%†46.8%†
Arena Hard PromptsArena (LMArena)15151471
LiveBench 1.0 ReasoningLiveBench91.2%–
GPQA Diamond (AA)Artificial Analysis93.2%87.9%
Humanity's Last Exam (AA)Artificial Analysis54.9%30.7%
GPQA Diamond (Vals)Vals AI93.4%†84.1%
Kagi LLM BenchmarkKagi LLM Benchmark–78.5%
ARC-AGI-1ARC Prize97.5%65.3%†
ARC-AGI-2ARC Prize90.4%11.8%†
ARC-AGI-3ARC Prize30.2%†–
CritPtArtificial Analysis29.1%3.1%
MysteryMechanismVals AI37.4%†–
Chess PuzzlesEpoch AI Benchmarking Hub43.0%–
Mystery Game PuzzlesEpoch AI Benchmarking Hub59.0%–
CL-benchEpoch AI Benchmarking Hub–19.3%†
CL-bench LifeEpoch AI Benchmarking Hub–13.2%†
LMCAEpoch AI Benchmarking Hub63.3%–
DTBenchEpoch AI Benchmarking Hub97.6%–
ForecastBenchForecastBench66.5%–
BTF-3Epoch AI Benchmarking Hub12.0%†–
Coding
SWE-bench Verified (Epoch)Epoch AI Benchmarking Hub–73.8%†
SciCodeSciCode56.4%49.0%†
WeirdMLWeirdML91.8%45.6%†
Show all 146 benchmarks
MetricPublisherClaude Opus 5 (max)Kimi K2.5 (thinking)
Coding
FrontierCodeCognition53.4%–
Arena CodingArena (LMArena)15281502
Arena WebDevArena (LMArena)16901437
LiveBench 1.0 CodingLiveBench81.5%–
SciCode (AA)Artificial Analysis56.4%–
LiveCodeBenchVals AI89.0%†83.9%
IOIVals AI84.3%†–
SWE-bench (Vals)Vals AI97.0%†70.0%
SWE-bench Verified (bash only)SWE-bench–70.8%†
SWE-bench Verified (any scaffold)SWE-bench–70.8%†
Code MigrationVals AI57.5%†7.0%
ProgramBenchVals AI3.0%†–
Vibe Code Bench 1-100Vals AI28.5%†–
Vibe Code Bench v1.1Vals AI88.4%†17.5%
SWE Atlas: Codebase QnAScale AI SEAL63.2%†13.1%†
SWE Atlas: RefactoringScale AI SEAL–20.9%†
SWE Atlas: Test WritingScale AI SEAL62.2%†25.8%†
DeepSWE v1.1Epoch AI Benchmarking Hub73.7%–
CursorBenchCursor46.6%–
ALE-BenchEpoch AI Benchmarking Hub2164.6†821.6†
FrontierSWEEpoch AI Benchmarking Hub52.0%–
Terminal-Bench 4.0 (AA)Artificial Analysis49.0%–
Terminal-Bench 2.1 (AA)Artificial Analysis89.1%45.7%
Startrise FrontendStartrise82.3%†–
Agents & tools
Terminal-BenchTerminal-Bench51.8%43.2%†
OSWorld-Verified 2.0OSWorld31.4%–
APEX-AgentsMercor65.8%–
Arena AgentArena (LMArena)8.2–
LiveBench Agentic CodingLiveBench65.2%–
Terminal-Bench HardArtificial Analysis–34.9%
τ²-Bench Telecom (AA)Artificial Analysis–95.9%
Terminal-Bench 2.1 (Vals)Vals AI84.6%†42.0%
MCP AtlasScale AI SEAL85.8%†64.4%†
HiL-BenchScale AI SEAL57.0%†–
τ³-bench Bankingτ²-bench48.7%–
τ³-Banking (AA)Artificial Analysis42.1%14.2%
Analyst Agent (AA)Artificial Analysis53.8%–
APEX-Agents (AA)Artificial Analysis–11.5%
CUA-benchVals AI9.0%–
CyberBenchVals AI65.4%†–
Finance Agent v2Vals AI58.6%†35.8%
Harvey's Legal Agent Benchmark v1Vals AI6.7%†0.0%
Legal Research BenchVals AI55.3%†15.9%
SkillsBenchVals AI60.4%†–
SREBenchVals AI12.2%†–
Tax Agent BenchVals AI45.5%†–
Terminal-Bench 2.0 (Vals)Vals AI–40.5%
Terminal-Bench 4.0 (Vals)Vals AI53.5%†–
Terminal-Bench ScienceVals AI27.1%†–
Time Horizon Index: KSPVals AI18.8%†–
Arena SearchArena (LMArena)1214†–
BALROGBALROG63.4%–
PostTrainBenchEpoch AI Benchmarking Hub35.0%†–
GBAEvalEpoch AI Benchmarking Hub79.6%†–
Vending-Bench 2Andon Labs11181.9†1198.5†
AutomationBenchArtificial Analysis56.6%–
GDP.pdf (AA)Artificial Analysis21.6%–
EnterpriseOps-GymArtificial Analysis47.5%–
AA-Briefcase v1.1Artificial Analysis1655–
Terminal-Bench-Science 0.1 (AA)Artificial Analysis28.6%–
DeepSec-Bench (AA)Artificial Analysis25.9%–
FACTS SearchGoogle85.3%†–
Maths
FrontierMath Tiers 1–3Epoch AI Benchmarking Hub85.6%–
FrontierMath Tier 4Epoch AI Benchmarking Hub73.2%–
OTIS Mock AIMEEpoch AI Benchmarking Hub98.9%92.2%†
ProofBenchVals AI99.0%–
LiveBench MathematicsLiveBench95.7%–
AIME (Vals)Vals AI–95.6%
AIME 2026MathArena–95.8%
HMMT February 2026MathArena–87.1%
MathArena ApexMathArena–8.8%
Arena MathsArena (LMArena)15171471
Knowledge
SimpleQA VerifiedEpoch AI Benchmarking Hub59.9%34.3%†
LiveBench 1.0 Data AnalysisLiveBench74.5%–
MMLU-ProVals AI91.6%†85.9%
LegalBenchVals AI87.0%†–
CorpFinVals AI73.2%†68.3%
TaxEvalVals AI75.1%†74.2%
MedQAVals AI–94.4%
PRBench FinanceScale AI SEAL–46.5%†
PRBench LegalScale AI SEAL–43.8%†
MultiNRCScale AI SEAL–35.2%†
BioMysteryBenchVals AI79.3%†–
CaseLaw v2Vals AI–58.7%
Excel Modeling BenchmarkVals AI73.6%†28.5%
MedCodeVals AI63.6%†39.3%
MedScribeVals AI91.0%†76.4%
Public Benefits Bench v1.1Vals AI76.9%†–
DrugDiscoveryBenchScale AI SEAL53.7%–
EBR-benchEpoch AI Benchmarking Hub45.7%–
GDP.pdfEpoch AI Benchmarking Hub24.0%–
Instruction following
LiveBench LanguageLiveBench88.7%–
LiveBench Instruction FollowingLiveBench63.8%–
IFBenchArtificial Analysis–70.2%
MultiChallengeScale AI SEAL–61.4%†
TutorBenchScale AI SEAL–54.6%†
Arena Instruction FollowingArena (LMArena)14951439
Human preference
Arena TextArena (LMArena)14911450
EQ-Bench 4EQ-Bench1365†–
Arena Creative WritingArena (LMArena)14741423
Arena Multi-turnArena (LMArena)14871452
Arena Longer QueriesArena (LMArena)15031458
Multimodal
Arena VisionArena (LMArena)1320†1269
MMMU-ProArtificial Analysis84.7%75.4%
VISTAScale AI SEAL–41.9%†
MMMU-Pro (Vals)Vals AI89.9%†84.3%
MortgageTaxVals AI72.1%†66.5%
Blueprint-Bench 2Epoch AI Benchmarking Hub30.4%†–
Long context
Fiction.LiveBench 120kFiction.live–78.1%†
AA-LCR v1.1Artificial Analysis79.3%78.0%
Arena DocumentArena (LMArena)1490†1436
MLCRArtificial Analysis11.1%–
Composite
Epoch Capabilities IndexEpoch AI Benchmarking Hub163.9†148.1†
LiveBench 1.0LiveBench80.1%–
AA Intelligence Index v4.3.2Artificial Analysis50.823.5
AA Progress Index v1Artificial Analysis333.7270.5
Vals RSI IndexVals AI33.0%6.8%†
Vals Science IndexVals AI47.9%†–
Vals Multimodal IndexVals AI73.9%†–
Safety & honesty
FORTRESSScale AI SEAL16.8%†41.1%†
MASKScale AI SEAL–70.5%†

Data as of 2026-10-29. Best configuration of each model; every score links to its source on the model pages.

What changed

  • Claude Opus 5 (max): BenchLeader Index 66.3 → 65.5.
  • Claude Opus 5 (max): 1 new result (AA Progress Index v1).
  • Kimi K2.5 (thinking): 1 new result (AA Progress Index v1).
  • Claude Opus 5 (max): BenchLeader Index 66.8 → 66.3.
  • Claude Opus 5 (max): 1 new result (Vals RSI Index).
  • Kimi K2.5 (thinking): 1 new result (Harvey's Legal Agent Benchmark v1).
  • Claude Opus 5 (max): 1 score revision.
  • Kimi K2.5 (thinking): output price moved from $2.75/M to $3.00/M (9% up).
  • Claude Opus 5 (max): 1 new result (ForecastBench); 1 score revision.
  • Claude Opus 5 (max): 4 new results (GDPval-AA v2.1, normalised score, AA-Omniscience v1.1: accuracy, AA-Omniscience v1.1: non-hallucination…).

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

Claude Opus 5 vs Kimi K2.5: questions

Is Claude Opus 5 better than Kimi K2.5?
Claude Opus 5 (max) leads on quality: 65.5 vs 58.8. The BenchLeader Index combines every independent quality benchmark; Claude Opus 5 (max) is ahead overall as of 29 Oct 2026, but check the category scores for your use.
Claude Opus 5 vs Kimi K2.5: which is cheaper, and will that change?
Kimi K2.5 is cheaper today: $1.20 against $10.00 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
Is Claude Opus 5 or Kimi K2.5 better for coding?
Claude Opus 5 scores higher in coding (70 vs 56 on the category index, where 50 is average). On Arena WebDev, the top shared coding benchmark, it is 1690 for Claude Opus 5 against 1437 for Kimi K2.5.
Which is faster, Claude Opus 5 or Kimi K2.5?
Kimi K2.5 streams faster (87 vs 56 output tokens per second, Artificial Analysis 7-day median).
Is Claude Opus 5 or Kimi K2.5 better for agents and tool use?
Claude Opus 5 scores higher in agentic tasks (64 vs 45 on the category index, where 50 is average). On τ³-Banking (AA), the top shared agents & tools benchmark, it is 42.1% for Claude Opus 5 against 14.2% for Kimi K2.5.
More questions (4)
Which is better for long documents?
They are level on long-context benchmarks (65 vs 64). Claude Opus 5 accepts more context: 1,000,000 against 262,144 tokens for Kimi K2.5.
Which has lower latency?
Kimi K2.5 answers first: 36.0 s against 43.9 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 vs Kimi K2.5?
Across the 16 benchmarks both have, Claude Opus 5 leads 16, Kimi K2.5 leads 0. Every overall index that rates both puts Claude Opus 5 ahead.
When was Claude Opus 5 released?
Claude Opus 5 was released on 24 Jul 2026.