Skip to main content

Model comparison

Claude Opus 5 vs Kimi K3

Data verified

Head-to-head evidence from 27 shared benchmark results across 5 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.

Anthropic
85.88/100
Margin
5.9pts
← winning
Moonshot AI
79.98/100
2 category wins1 category wins

Public leaderboard positions: Claude Opus 5 #1 (Estimated); Kimi K3 #5 (Supported). Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.

Evidence parity. Claude Opus 5 and Kimi K3 share 27 comparable benchmark results. 3 of 8 categories are comparable. 52 results are unique to Claude Opus 5; 30 to Kimi K3.

Updated July 24, 2026
Shared results
27
Claude Opus 5 only
52
Kimi K3 only
30
Comparable categories
3 / 8

Pick Claude Opus 5 if you want the stronger benchmark profile. Kimi K3 only becomes the better choice if multimodal & grounded is the priority or you want the cheaper token bill.

Confidence note. This is a partial-evidence comparison with 27 shared benchmark results across 5 evidence categories; 3 of 8 categories currently have scoreable aggregates for both models. Treat the verdict as directional until coverage is more balanced.

Why this result

Claude Opus 5 is clearly ahead on the BenchAlign aggregate, 85.88 to 79.98. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.

Claude Opus 5's sharpest advantage is in knowledge, where it averages 64.7 against 61. The single biggest benchmark swing on the page is HLE, 64.7% to 56%. Kimi K3 does hit back in multimodal & grounded, so the answer changes if that is the part of the workload you care about most.

Claude Opus 5 is also the more expensive model on tokens at $5.00 input / $25.00 output per 1M tokens, versus $3.00 input / $15.00 output per 1M tokens for Kimi K3.

Category breakdown

Exact category averages are shown below. Not measured means BenchLM does not have enough sourced public coverage for that model and category.

Category scores and score margins for Claude Opus 5 and Kimi K3
CategoryClaude Opus 5ΔKimi K3
MultimodalClaude Opus 566.9Margin 11.6Kimi K378.5
KnowledgeClaude Opus 564.7Margin 3.7Kimi K361.0
AgenticClaude Opus 590.8Margin 1.3Kimi K389.5
CodingClaude Opus 589.5MarginNo overlapKimi K3Not measured
ReasoningClaude Opus 590.4MarginNo overlapKimi K3Not measured

Decisive benchmark drivers

The largest measured benchmark gaps in this matchup, with exact reported values.

More
A · Claude Opus 5B · Kimi K3
  1. HLE

    Knowledge
    Source ↗
    A 64.7%B 56%
    Winner: Claude Opus 5Δ 8.7
    HLE: Claude Opus 5 scored 64.7%; Kimi K3 scored 56%. Claude Opus 5 wins this benchmark.
  2. OfficeQA Pro

    Multimodal
    Source ↗
    A 66.9%B 63.3%
    Winner: Claude Opus 5Δ 3.6
    OfficeQA Pro: Claude Opus 5 scored 66.9%; Kimi K3 scored 63.3%. Claude Opus 5 wins this benchmark.
  3. BrowseComp

    Agentic
    Source ↗
    A 90.8%B 91.2%
    Winner: Kimi K3Δ 0.4
    BrowseComp: Claude Opus 5 scored 90.8%; Kimi K3 scored 91.2%. Kimi K3 wins this benchmark.

Operational comparison

Runtime and commercial metrics are compared only when both models have a complete sourced value.

MetricClaude Opus 5Kimi K3Comparison
Input / output priceUSD per 1M tokensClaude Opus 5$5 input / $25 outputKimi K3$3 input / $15 outputKimi K3 has the lower combined listed price.
Generation speedtokens per secondClaude Opus 5Not availableKimi K3Not availableA complete speed comparison is not available.
First-answer latencyseconds to first tokenClaude Opus 5Not availableKimi K3Not availableA complete latency comparison is not available.
Context windowmaximum listed tokensClaude Opus 5Not availableKimi K31.05MA complete context comparison is not available.

Benchmark Deep Dive

AgenticClaude Opus 5 wins
BenchmarkClaude Opus 5Kimi K3Result
FrontierBench v0.1Source 43.3%Not comparable
BrowseCompSource 90.8%91.2%Kimi K3 leads
HLE w/ toolsSource 64.7%Not comparable
DeepSearchQASource 95.0%95.0%Tie
DRACOSource 88.6%Not comparable
BrowseComp (10-agent, prerelease)Source 93.6%Not comparable
OSWorld 2.0Source 70.6%Not comparable
MCP AtlasSource 85.8%84.2%Claude Opus 5 leads
MCP-Atlas claim coverageSource 89.1%Not comparable
LAB all-pass (Anthropic harness)Source 23.58%Not comparable
LAB criterion-pass (Anthropic harness)Source 93.74%Not comparable
LAB all-pass (Harvey held-out)Source 11.7%Not comparable
LAB criterion-pass (Harvey held-out)Source 94.1%Not comparable
GDPval-AASource 18611686Claude Opus 5 leads
Toolathlon-VerifiedSource 80.6%73.2%Claude Opus 5 leads
Toolathlon Verified Pass@3Source 87.0%Not comparable
Toolathlon Verified Pass³Source 73.1%Not comparable
Toolathlon Verified avg. turnsSource 23.5 turnsNot comparable
AutomationBenchSource 26.0%30.8%Kimi K3 leads
AA Agentic IndexSource 55.3%50.1%Claude Opus 5 leads
GDPval-AASource 68.0%59.3%Claude Opus 5 leads
AA Tau3 BankingSource 30.3%33.4%Kimi K3 leads
AA BriefcaseSource 17201541Claude Opus 5 leads
aaTerminalBench21Source 89.1%85%Claude Opus 5 leads
Terminal-Bench 2.0Source 88.3%Not comparable
JobBenchSource 52.9%Not comparable
APEX-AgentsSource 37.6%Not comparable
SpreadsheetBench 2Source 34.8%Not comparable
DECK-BenchSource 73.5%Not comparable
AA AutomationBenchSource 52.7%Not comparable
AA EnterpriseOps-GymSource 45.3%Not comparable
AA Harvey LABSource 94.6%Not comparable
APEX-Agents-AASource 41.3%Not comparable
AA ITBenchSource 47.7%Not comparable
Coding
BenchmarkClaude Opus 5Kimi K3Result
SWE-bench VerifiedSource 96%Not comparable
SWE-bench ProSource 79.2%Not comparable
SWE MultilingualSource 89.5%Not comparable
SWE MultimodalSource 59.4%Not comparable
deepSweSource 68.8%67.5%Claude Opus 5 leads
FrontierCode 1.1 MainSource 53.4%Not comparable
FrontierCode 1.1 ExtendedSource 63.6%Not comparable
ProgramBench (episode 1)Source 83.0%Not comparable
ProgramBenchSource 93.0%77.8%Claude Opus 5 leads
cursorBench32Source 70.0%Not comparable
AA Coding IndexSource 78.0%76.2%Claude Opus 5 leads
AA-SciCodeSource 55.7%58.7%Kimi K3 leads
FrontierSWESource 81.2%Not comparable
Kimi Code Bench v2Source 72.9%Not comparable
sweMarathonSource 42%Not comparable
PostTrain BenchSource 36.6%Not comparable
MLS-Bench LiteSource 48.3%Not comparable
Reasoning
BenchmarkClaude Opus 5Kimi K3Result
ARC-AGI-1Source 97.50%Not comparable
ARC-AGI-2Source 90.4%Not comparable
ARC-AGI-3Source 30.2%Not comparable
AA-LCRSource 70.0%74.7%Kimi K3 leads
CritPtSource 29.1%23.4%Claude Opus 5 leads
KnowledgeClaude Opus 5 wins
BenchmarkClaude Opus 5Kimi K3Result
HLESource 64.7%56%Claude Opus 5 leads
HLE w/o toolsSource 56.3%43.5%Claude Opus 5 leads
HealthBench (raw)Source 67.1%Not comparable
HealthBench (length-adjusted)Source 57.8%Not comparable
HealthBench ProfessionalSource 59.8%Not comparable
HealthBench Professional (raw)Source 73.4%Not comparable
BioMysteryBench (human-solvable)Source 90.1%Not comparable
BioMysteryBench (human-difficult)Source 49.4%Not comparable
SpatialBench VerifiedSource 72.5%Not comparable
SingleCellBenchSource 60.6%Not comparable
ProteinGym HardSource 47.7%Not comparable
Protein DesignSource 42.5%Not comparable
Organic chemistry V2Source 61.6%Not comparable
Protocols (troubleshooting)Source 61.1%Not comparable
Protocols (understanding)Source 78.4%Not comparable
Artificial Analysis Intelligence IndexSource 60.7%57.1%Claude Opus 5 leads
AA-GPQA DiamondSource 93.2%93.5%Kimi K3 leads
AA-HLESource 52.6%44.3%Claude Opus 5 leads
AA-Omniscience IndexSource 31.3%18.4%Claude Opus 5 leads
AA-Omniscience AccuracySource 54.2%46.0%Claude Opus 5 leads
AA-Omniscience Hallucination RateSource 50.1%50.9%Claude Opus 5 leads
GPQASource 93.5%Not comparable
GPQA-DSource 93.5%Not comparable
Math
BenchmarkClaude Opus 5Kimi K3Result
IMO 2026Source 42/42Not comparable
RiemannBench (no tools)Source 60.0%Not comparable
RiemannBench (tools)Source 79.0%Not comparable
ArXivMath Jun. 2026 (no tools)Source 90.8%Not comparable
ArXivMath Jun. 2026 (tools)Source 91.3%Not comparable
Multilingual
BenchmarkClaude Opus 5Kimi K3Result
GMMLUSource 92.5%Not comparable
MILUSource 92.1%Not comparable
INCLUDESource 89.8%Not comparable
MultimodalKimi K3 wins
BenchmarkClaude Opus 5Kimi K3Result
Chartography (no tools)Source 29.6%Not comparable
Chartography (tools)Source 83.0%Not comparable
BenchCAD Vision2Code (no tools)Source 0.366Not comparable
BenchCAD Vision2Code (tools)Source 0.821Not comparable
GDP.pdf (no tools)Source 83.4%Not comparable
GDP.pdf (tools)Source 85.5%Not comparable
OfficeQASource 78.1%Not comparable
OfficeQA ProSource 66.9%63.3%Claude Opus 5 leads
AA-MMMU-ProSource 84.7%80.5%Claude Opus 5 leads
MMMU-ProSource 81.6%Not comparable
MMMU-Pro w/ PythonSource 83.4%Not comparable
CharXiv w/o toolsSource 84.8%Not comparable
CharXivSource 91.3%Not comparable
MathVisionSource 94.3%Not comparable
MathVision w/ PythonSource 97.8%Not comparable
BabyVision w/ PythonSource 85.7%Not comparable
ZeroBenchSource 23.0%Not comparable
ZeroBench w/ PythonSource 41.0%Not comparable
WorldVQA ForceAnswerSource 51.0%Not comparable
OmniDocBenchSource 91.1%Not comparable
PerceptionBenchSource 58.5%Not comparable
Design Arena WebsiteSource 1386Not comparable
Frequently Asked Questions (4)

Which is better, Claude Opus 5 or Kimi K3?

Claude Opus 5 is ahead on BenchLM's BenchAlign leaderboard, 85.88 to 79.98. The biggest single separator in this matchup is HLE, where the scores are 64.7% and 56%.

Which is better for knowledge tasks, Claude Opus 5 or Kimi K3?

Claude Opus 5 has the edge for knowledge tasks in this comparison, averaging 64.7 versus 61. Inside this category, AA-Omniscience Index is the benchmark that creates the most daylight between them.

Which is better for agentic tasks, Claude Opus 5 or Kimi K3?

Claude Opus 5 has the edge for agentic tasks in this comparison, averaging 90.8 versus 89.5. Inside this category, AA Briefcase is the benchmark that creates the most daylight between them.

Which is better for multimodal and grounded tasks, Claude Opus 5 or Kimi K3?

Kimi K3 has the edge for multimodal and grounded tasks in this comparison, averaging 78.5 versus 66.9. Inside this category, AA-MMMU-Pro is the benchmark that creates the most daylight between them.

Related Comparisons

Last updated: July 24, 2026