BenchLeader

Claude Sonnet 5.5 vs Kimi K2.5

Full profiles: Claude Sonnet 5.5 (Anthropic) and Kimi K2.5 (Moonshot AI): every benchmark, price and rank.

Updated · Data from 14 sources

As of 16 Nov 2026, Claude Sonnet 5.5 is ahead of Kimi K2.5 on quality (64.2 vs 59.4 on the BenchLeader Index), but Kimi K2.5 is 3.3× cheaper at $1.20 vs $4.00 per 1M tokens blended. Choose Claude Sonnet 5.5 for agents & tools (71 vs 45), coding (70 vs 55), and maths (71 vs 65), where it also streams 3×+ faster; choose Kimi K2.5 for fast first answers (42.0 s vs 290.5 s) and knowledge (60 vs 57).

Key facts (as of 16 Nov 2026)

  • Claude Sonnet 5.5 scores 64.2 on the BenchLeader Index and Kimi K2.5 scores 59.4 (BenchLeader, 16 Nov 2026).
  • Claude Sonnet 5.5 costs $4.00 per 1M tokens blended (list price, 16 Nov 2026).
  • Kimi K2.5 costs $1.20 per 1M tokens blended (list price, 16 Nov 2026).
  • Claude Sonnet 5.5 streams 3×+ faster than Kimi K2.5 (Artificial Analysis, 7-day median, 16 Nov 2026).
  • Of the 13 benchmarks both models have, Claude Sonnet 5.5 leads 13 (BenchLeader, 16 Nov 2026).

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 Sonnet 5.570 vs 55 category score; Code Migration 69.8% vs 7.0%High
Agents & tool useClaude Sonnet 5.571 vs 45 category score; GDPval-AA v2.1 1838 vs 847High
MathsClaude Sonnet 5.571 vs 65 category scoreMedium
ReasoningEither68 vs 67 category score; Humanity's Last Exam (AA) 55.0% vs 30.7%Either
KnowledgeKimi K2.560 vs 57 category score; Excel Modeling Benchmark 28.5% vs 75.7%Medium
Long documentsClaude Sonnet 5.566 vs 64 category score; AA-LCR v1.1 82.7% vs 78.0%Medium
Chat quality (human preference)–Only Kimi K2.5 has an Arena rating (63 category score)Not comparable
Lowest costKimi K2.5$1.20 vs $4.00 per 1M blended; agentic coding session $0.048 vs $0.160–
Fastest first answerKimi K2.542.0 s (Kimi K2.5 (thinking)) vs 290.5 s (Claude Sonnet 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 Sonnet 5.5 (max) leads on quality: 64.2 vs 59.4.
  • Claude Sonnet 5.5 (max) is stronger in agents & tools, coding, long context, maths, reasoning.
  • Kimi K2.5 (thinking) is stronger in knowledge.
  • Kimi K2.5 (thinking) is 3.3× cheaper ($1.20 vs $4.00 per 1M blended).
  • Claude Sonnet 5.5 (max) streams 3.6× faster (152 vs 43 tokens per second).

At a glance

SpecClaude Sonnet 5.5Kimi K2.5
DeveloperAnthropicMoonshot AI
Released28 Sep 202627 Jan 2026
Context window (tokens)1,000,000262,144
Compared atClaude Sonnet 5.5 (max)Kimi K2.5 (thinking)
Open weightsNoYes
Price in / out / cached input per 1M$2.00 / $10.00 / $0.100$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 Sonnet 5.5 (max)Kimi K2.5 (thinking)Leader
Reasoning67.867.3level
Coding70.255.4Claude Sonnet 5.5
Agents & tools70.545.1Claude Sonnet 5.5
Maths71.365.1Claude Sonnet 5.5
Knowledge57.460.1Kimi K2.5
Instruction following–69.1not comparable
Human preference–63.2not comparable
Multimodal–59.1not comparable
Long context66.063.6Claude Sonnet 5.5
Composite63.0–not comparable

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 contextClaude Sonnet 5.5, Reasoning: 67.8Claude Sonnet 5.5, Coding: 70.2Claude Sonnet 5.5, Agents & tools: 70.5Claude Sonnet 5.5, Maths: 71.3Claude Sonnet 5.5, Knowledge: 57.4Claude Sonnet 5.5, Long context: 66.0Kimi K2.5, Reasoning: 67.3Kimi K2.5, Coding: 55.4Kimi K2.5, Agents & tools: 45.1Kimi K2.5, Maths: 65.1Kimi K2.5, Knowledge: 60.1Kimi K2.5, Long context: 63.6

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

Model, click to add or removeIndex

Where sources disagree

Across the 13 benchmarks both have, Claude Sonnet 5.5 leads 13, Kimi K2.5 leads 0. Every overall index that rates both puts Claude Sonnet 5.5 ahead.

Overall indexClaude Sonnet 5.5 (max)Kimi K2.5 (thinking)Leader
BenchLeader Index64.259.4Claude Sonnet 5.5
Arena Text1473†1450Claude Sonnet 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 Sonnet 5.5 (max)Kimi K2.5 (thinking)
Blended price per 1M tokens$4.00$1.20
Input / output per 1M$2.00 / $10.00$0.600 / $3.00
Cached input per 1M$0.100$0.100
Price per task: chat reply$0.0038$0.0011
Price per task: summarize a 30-page report$0.030$0.0090
Price per task: code edit$0.027$0.0081
Price per task: agentic coding session$0.160$0.048
Price per task: structured extraction$0.0060$0.0018
Agentic coding session, input mostly cached$0.074$0.025
Cost per benchmark task (AA Index)$10.02–
Output tokens per task (AA Index)285k–
Of which reasoning206k–
Time per task (AA Index)31 min–
Output speed152 tok/s43 tok/s
Time to first answer290.5 s42.0 s
Time to a complete answer, last measured 13 Nov 2026487.7 s129.1 s

Each model is compared at its best-ranked reasoning-effort setting. Claude Sonnet 5.5 (max) scores 64.2 and takes 291 s to its first answer token; Claude Sonnet 5.5 (xhigh) scores 62.5 in 27 s, Claude Sonnet 5.5 (high) scores 59.5 in 9 s. Kimi K2.5 (thinking) scores 59.4 and takes 42 s to its first answer token; Kimi K2.5 scores 54.6 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

190 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 Sonnet 5.5 (max)Kimi K2.5 (thinking)
BenchLeader IndexBenchLeader64.259.4
Agents & tools scoreBenchLeader70.545.1
Coding scoreBenchLeader70.255.4
Composite scoreBenchLeader63.0–
Knowledge scoreBenchLeader57.460.1
Long context scoreBenchLeader66.063.6
Maths scoreBenchLeader71.365.1
Reasoning scoreBenchLeader67.867.3
Human preference scoreBenchLeader–63.2
Instruction following scoreBenchLeader–69.1
Multimodal scoreBenchLeader–59.1
Blended price $/MBenchLeader$4.00$1.20
Output speedBenchLeader152 tok/s43 tok/s
Time to first answerBenchLeader290.5 s42.0 s
Context windowBenchLeader1,000,000 tokens262,144 tokens
Reasoning
GPQA DiamondEpoch AI Benchmarking Hub95.6%87.6%†
Humanity's Last ExamScale AI / CAIS46.1%†24.4%†
SimpleBenchSimpleBench75.9%†46.8%†
Arena Hard PromptsArena (LMArena)1499†1471
LiveBench 2.0 ReasoningLiveBench91.6%–
LiveBench 1.0 ReasoningLiveBench91.6%–
GPQA Diamond (AA)Artificial Analysis–87.9%
Humanity's Last Exam (AA)Artificial Analysis55.0%30.7%
HLE-Diamond (AA)Artificial Analysis53.3%–
GPQA Diamond (Vals)Vals AI–84.1%
Kagi LLM BenchmarkKagi LLM Benchmark–78.5%
ARC-AGI-1ARC Prize96.5%65.3%†
ARC-AGI-2ARC Prize83.9%†11.8%†
ARC-AGI-3ARC Prize72.4%†–
CritPtArtificial Analysis31.4%3.1%
CritPt v2 hard subset (AA)Artificial Analysis59.0%–
MysteryMechanismVals AI49.1%–
HLE-Diamond (Scale)Scale AI SEAL37.5%†–
Arena ExpertArena (LMArena)1512†1482
Chess PuzzlesEpoch AI Benchmarking Hub59.0%–
Mystery Game PuzzlesEpoch AI Benchmarking Hub65.0%–
CL-benchEpoch AI Benchmarking Hub–19.3%†
CL-bench LifeEpoch AI Benchmarking Hub–13.2%†
LMCAEpoch AI Benchmarking Hub68.3%–
DTBenchEpoch AI Benchmarking Hub97.3%–
Show all 190 benchmarks
MetricPublisherClaude Sonnet 5.5 (max)Kimi K2.5 (thinking)
Coding
SWE-bench Verified (Epoch)Epoch AI Benchmarking Hub–73.8%†
SciCodeSciCode61.0%49.0%†
WeirdMLWeirdML–45.6%†
FrontierCodeCognition52.1%†–
Arena CodingArena (LMArena)1517†1502
Arena WebDevArena (LMArena)1772†1436
LiveBench 2.0 CodingLiveBench77.2%–
LiveBench 1.0 CodingLiveBench91.4%–
SciCode (AA)Artificial Analysis61.0%–
LiveCodeBenchVals AI–83.9%
IOIVals AI83.1%–
SWE-bench (Vals)Vals AI–70.0%
SWE-bench Verified (bash only)SWE-bench–70.8%†
SWE-bench Multilingual (bash only)SWE-bench–67.3%†
SWE-bench Verified (any scaffold)SWE-bench–70.8%†
Code MigrationVals AI69.8%7.0%
Code Migration v1.1Vals AI86.9%–
ProgramBenchVals AI6.5%–
SWE-bench MultimodalVals AI58.3%–
Vibe Code Bench v1.1Vals AI92.4%17.5%
SWE Atlas v1.0: Codebase QnAScale AI SEAL–13.1%†
SWE Atlas v1.0: RefactoringScale AI SEAL–20.9%†
SWE Atlas v1.0: Test WritingScale AI SEAL–25.8%†
SWE Atlas v1.1Scale AI SEAL68.3%†–
SWE Atlas v1.1: Codebase QnAScale AI SEAL74.2%†–
SWE Atlas v1.1: Test WritingScale AI SEAL67.8%†–
SWE Atlas v1.1: RefactoringScale AI SEAL62.9%†–
SWE-Bench Pro V2Scale AI SEAL95.8%†–
SWE-Bench Pro V2 HARDScale AI SEAL70.6%†–
MirrorCodeEpoch AI Benchmarking Hub97.8%–
CursorBenchCursor55.5%–
ALE-BenchEpoch AI Benchmarking Hub1819.1†821.6†
FrontierSWEEpoch AI Benchmarking Hub61.9%–
WeirdML v3WeirdML19.9%†–
Terminal-Bench 4.0 (AA)Artificial Analysis63.6%–
Terminal-Bench 2.1 (AA)Artificial Analysis–45.7%
Startrise FrontendStartrise75.9%†–
GameBench 2 OverallDuelLab GameBench 260.4†–
GameBench 2 by reasoning settingDuelLab GameBench 291.4†–
Design Arena WebsiteDesign Arena1309†1241
FrontierCode 1.1 (AA)Artificial Analysis45.8%–
FrontierCode 2.0 (AA)Artificial Analysis41.7%–
Agents & tools
Terminal-BenchTerminal-Bench63.9%43.2%†
APEX-AgentsMercor75.5%–
Arena AgentArena (LMArena)16–
LiveBench Long AgenticLiveBench63.4%–
LiveBench BrowserLiveBench89.3%–
LiveBench Agentic CodingLiveBench56.3%–
Terminal-Bench HardArtificial Analysis–34.9%
τ²-Bench Telecom (AA)Artificial Analysis–95.9%
Terminal-Bench 2.1 (Vals)Vals AI83.2%†42.0%
MCP AtlasScale AI SEAL–64.4%†
GDPval-AA v2.1, normalised scoreArtificial Analysis66.9%–
GDPval-AA v2.1Artificial Analysis1838847
τ³-Banking (AA)Artificial Analysis–14.2%
Analyst Agent (AA)Artificial Analysis57.5%–
APEX-Agents (AA)Artificial Analysis–11.5%
CyberBenchVals AI59.6%–
Finance Agent v2Vals AI58.1%35.8%
Finance Agent v2.1Vals AI67.3%–
Harvey's Legal Agent Benchmark v1.1Vals AI10.4%–
Harvey's Legal Agent Benchmark v1Vals AI2.9%0.0%
Legal Research BenchVals AI48.1%15.9%
SREBenchVals AI30.1%–
Tax Agent BenchVals AI42.0%–
Terminal-Bench 2.0 (Vals)Vals AI–40.5%
Terminal-Bench 4.0 (Vals)Vals AI64.1%–
Terminal-Bench ScienceVals AI47.1%–
Arena Agent Task CompletionArena (LMArena)19.5–
Vending-Bench 2Andon Labs11165.3†1198.5†
SpireClimbEpoch AI Benchmarking Hub1.3%–
AutomationBenchArtificial Analysis71.8%–
GDP.pdf (AA)Artificial Analysis25.8%–
Harvey LAB-AA v1.1Harvey2.8%–
AA-Briefcase v1.1Artificial Analysis1816–
GDPxGDPx by SemiAnalysis8.2%–
Design Arena AgenticDesign Arena1293†1116
Terminal-Bench-Science 0.1 (AA)Artificial Analysis53.3%–
MKA (AA)Artificial Analysis52.0%–
FACTS SearchGoogle86.8%†–
Maths
FrontierMath Tiers 1–3Epoch AI Benchmarking Hub88.8%–
FrontierMath Tier 4Epoch AI Benchmarking Hub80.5%–
OTIS Mock AIMEEpoch AI Benchmarking Hub100.0%92.2%†
ProofBenchVals AI100.0%–
LiveBench MathematicsLiveBench96.1%–
AIME (Vals)Vals AI–95.6%
AIME 2026MathArena–95.8%†
HMMT February 2026MathArena–87.1%†
MathArena ApexMathArena–8.8%†
Arena MathsArena (LMArena)1486†1470
FrontierMath ErdősEpoch AI Benchmarking Hub2.9%–
Knowledge
SimpleQA VerifiedEpoch AI Benchmarking Hub46.5%34.3%†
LiveBench 2.0 Data AnalysisLiveBench59.5%–
LiveBench ResearchLiveBench88.0%–
LiveBench 1.0 Data AnalysisLiveBench59.5%–
AA-Omniscience v1.1Artificial Analysis30.5–
MMLU-ProVals AI–85.9%
CorpFinVals AI–68.3%
TaxEvalVals AI–74.2%
MedQAVals AI–94.4%
PRBench FinanceScale AI SEAL–46.5%†
PRBench LegalScale AI SEAL–43.8%†
MultiNRCScale AI SEAL–35.2%†
BioMysteryBenchVals AI81.1%–
CaseLaw v2Vals AI–58.7%
ClaimBenchVals AI100.0%–
Excel Modeling BenchmarkVals AI75.7%28.5%
Excel Modeling Benchmark v1.1Vals AI75.9%–
MedCodeVals AI52.9%39.3%
MedScribeVals AI91.1%76.4%
Public Benefits Bench v1.1Vals AI67.2%–
Access to Justice BenchmarkVals AI74.9%–
Financial Coaching BenchVals AI74.9%–
Immigration BenchVals AI69.0%–
GDP.pdfEpoch AI Benchmarking Hub31.4%–
AA-Omniscience v1.1: accuracyArtificial Analysis51.5%–
AA-Omniscience v1.1: non-hallucinationArtificial Analysis56.7%–
Global-MMLU-Lite (AA)Artificial Analysis–84.0%†
HealthBench Professional (AA)OpenAI63.0%–
FACTS ParametricGoogle35.1%†–
Instruction following
LiveBench LanguageLiveBench78.0%–
LiveBench Instruction FollowingLiveBench56.8%–
IFBenchArtificial Analysis–70.2%
MultiChallengeScale AI SEAL–61.4%†
TutorBenchScale AI SEAL–54.6%†
Arena Instruction FollowingArena (LMArena)1482†1439
Human preference
Arena TextArena (LMArena)1473†1450
EQ-Bench Creative Writing v3EQ-Bench–1579†
Arena Creative WritingArena (LMArena)1464†1422
Arena Multi-turnArena (LMArena)1469†1452
Arena Longer QueriesArena (LMArena)1491†1457
Multimodal
Arena VisionArena (LMArena)1284†1269
LiveBench DesignLiveBench81.3%–
MMMU-ProArtificial Analysis–75.4%
VISTAScale AI SEAL–41.9%†
MMMU-Pro (Vals)Vals AI–84.3%
MortgageTaxVals AI–66.5%
SAGEVals AI88.0%–
Blueprint-Bench 2Epoch AI Benchmarking Hub39.3%†–
Furniture AssemblyEpoch AI Benchmarking Hub75.0%–
VIALS-AAHandshake AI59.6%–
Code2Video-AAArtificial Analysis1443–
AA-3DArtificial Analysis1204–
FACTS MultimodalGoogle50.4%†–
Long context
Fiction.LiveBench 120kFiction.live–78.1%†
AA-LCR v1.1Artificial Analysis82.7%78.0%
Arena DocumentArena (LMArena)–1438
MLCRArtificial Analysis41.7%–
FACTS GroundingGoogle62.5%†–
Composite
Epoch Capabilities IndexEpoch AI Benchmarking Hub166.6†148.8†
LiveBench 2.0LiveBench77.0%–
LiveBench 1.0LiveBench75.7%–
AA Intelligence Index v5.0.1Artificial Analysis54.5–
AA Intelligence Index v4.3.2Artificial Analysis5623.5
AA Progress Index v1Artificial Analysis350.5270.5
Vals IndexVals AI70.8–
Vals Finance IndexVals AI64.5%–
Vals RSI IndexVals AI–6.8%†
Vals Science IndexVals AI59.1%–
Vals Router IndexVals AI69.0%†–
Vals Legal IndexVals AI68.0%–
Safety & honesty
FORTRESSScale AI SEAL–41.1%†
MASKScale AI SEAL–70.5%†
DistressBenchScale AI SEAL74.6%†–
Metadata
AA Openness IndexArtificial Analysis–33.3

Data as of 2026-11-16. Best configuration of each model; every score links to its source on the model pages.

DuelLab GameBench 2, release gb2-20261111-7b598bb13de4 (https://benchmarks.duellab.org/releases/gb2-20261111-7b598bb13de4/), CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Scores mapped to BenchLeader model names; not endorsed by DuelLab.

What changed

  • Claude Sonnet 5.5 (max): 36 new results (LiveBench 2.0, LiveBench Long Agentic, LiveBench 2.0 Reasoning…).
  • Kimi K2.5 (thinking): 1 score revision.
  • Kimi K2.5 (thinking): 1 new result (GDPval-AA v2.1).
  • Claude Sonnet 5.5 (max): BenchLeader Index 65.1 → 64.6.
  • Kimi K2.5 (thinking): BenchLeader Index 58.9 → 59.4.
  • Claude Sonnet 5.5 (max): 1 new result (MirrorCode).
  • Claude Sonnet 5.5 (max): 2 score revisions.
  • Claude Sonnet 5.5 (max): 1 score revision.
  • Claude Sonnet 5.5 (max): 1 score revision.
  • Claude Sonnet 5.5 (max): 1 score revision.

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

Claude Sonnet 5.5 vs Kimi K2.5: questions

Is Claude Sonnet 5.5 better than Kimi K2.5?
Claude Sonnet 5.5 (max) leads on quality: 64.2 vs 59.4. The BenchLeader Index combines every independent quality benchmark; Claude Sonnet 5.5 (max) is ahead overall as of 16 Nov 2026, but check the category scores for your use.
Claude Sonnet 5.5 vs Kimi K2.5: which is cheaper, and will that change?
Kimi K2.5 is cheaper today: $1.20 against $4.00 per million tokens, blended at three input tokens per output token. No price change has been announced for either.
Is Claude Sonnet 5.5 or Kimi K2.5 better for coding?
Claude Sonnet 5.5 scores higher in coding (70 vs 55 on the category index, where 50 is average).
Which is faster, Claude Sonnet 5.5 or Kimi K2.5?
Claude Sonnet 5.5 streams 3×+ faster (152 vs 43 output tokens per second, Artificial Analysis 7-day median).
Is Claude Sonnet 5.5 or Kimi K2.5 better for agents and tool use?
Claude Sonnet 5.5 scores higher in agentic tasks (71 vs 45 on the category index, where 50 is average).
More questions (4)
Which is better for long documents?
Claude Sonnet 5.5 scores higher on long-context benchmarks (66 vs 64). Claude Sonnet 5.5 accepts more context: 1,000,000 against 262,144 tokens for Kimi K2.5.
Which has lower latency?
Kimi K2.5 answers first: 42.0 s against 290.5 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 Sonnet 5.5 vs Kimi K2.5?
Across the 13 benchmarks both have, Claude Sonnet 5.5 leads 13, Kimi K2.5 leads 0. Every overall index that rates both puts Claude Sonnet 5.5 ahead.
When was Claude Sonnet 5.5 released?
Claude Sonnet 5.5 was released on 28 Sep 2026.