Kimi K3 vs Claude Opus 5
Wins 2 of 8 areas
Long documents · Following instructions
Wins 5 of 8 areas
Coding · Agents · Reasoning · Facts · Math
Claude Opus 5 is the stronger all-rounder.Kimi K3 is cheaper and better at long documents.
Both rank among the ten best models we track in agents.
Scores updated · 43 tests both models report · How we compare
Where each one wins
Tests won in each of the eight areas we test. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
- ReasoningHard problems that need careful thinking04Claude Opus 54 of 6 tests · 2 ties
- MathCompetition and research-level math03Claude Opus 53 of 3 tests
- CodingWriting and fixing software13Claude Opus 53 of 5 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own23Claude Opus 53 of 6 tests · 1 tie
- FactsGetting facts right instead of making them up12Claude Opus 52 of 3 tests
- Long documentsFinding answers in very long texts10Kimi K31 of 1 test
- Following instructionsDoing exactly what it is asked10Kimi K31 of 1 test
- Images and chartsUnderstanding pictures, charts and video11Even1 each
Long documents and following instructions rest on a single test each.
What it costs
Prices per million tokens, roughly 750,000 words. The bars show the cost of a million tokens read plus a million written.
Kimi K3 costs 40% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where Claude Opus 5 pulls ahead
- Complex command-line tasks across many fieldsTerminal-Bench 4.0+36.4points ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+34.2points ahead
- Abstract visual puzzles that people can solveARC-AGI-2+30points ahead
Where Kimi K3 pulls ahead
- Reasons across sets of long documentsAA-LCR+9.4points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+7.6points ahead
- Reasoning about charts from research papersCharXiv (RQ)+7.6points ahead
Every test, side by side
All 43 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Opus 5
- Terminal-Bench 2.1Claude Opus 5 by 4.18589.1+4.1
- LMArena · WebDevClaude Opus 5 by 37 rating points16551692+37 rating
- SciCodeKimi K3 by 3.159.556.4+3.1
- LiveBench · Agentic CodingClaude Opus 5 by 362.265.2+3
- LiveBench · Codingtie81.581.5tie
AgentsClaude Opus 5
- Terminal-Bench 4.0Claude Opus 5 by 36.412.649+36.4
- GDPValClaude Opus 5 by 9.551.761.2+9.5
- τ-Bench V3 · BankingKimi K3 by 3.94642.1+3.9
- MCP AtlasClaude Opus 5 by 1.684.285.8+1.6
- OSWorld-VerifiedKimi K3 by 1.484.883.4+1.4
- BrowseComptie91.290.8tie
ReasoningClaude Opus 5
- ARC-AGI-2Claude Opus 5 by 3060.490.4+30
- SimpleBenchClaude Opus 5 by 19.960.780.6+19.9
- Humanity's Last ExamClaude Opus 5 by 846.954.9+8
- CritPtClaude Opus 5 by 5.723.429.1+5.7
- LiveBench · Reasoningtie90.791.2tie
- GPQA Diamondtie93.593.2tie
FactsClaude Opus 5
- AA-Omniscience · AccuracyClaude Opus 5 by 13.347.660.9+13.3
- SimpleQA VerifiedClaude Opus 5 by 9.350.659.9+9.3
- AA-Omniscience · Non-hallucinationKimi K3 by 7.646.839.2+7.6
Images and chartsEven
- CharXiv (RQ)Kimi K3 by 7.691.383.7+7.6
- MMMU-ProClaude Opus 5 by 4.280.584.7+4.2
MathClaude Opus 5
- FrontierMath Tier 4Claude Opus 5 by 34.23973.2+34.2
- FrontierMath Tiers 1-3 (v2)Claude Opus 5 by 13.472.285.6+13.4
- LiveBench · MathematicsClaude Opus 5 by 11.384.495.7+11.3
Following instructionsKimi K3
- LiveBench · Instruction FollowingKimi K3 by 7.671.463.8+7.6
Other results16 tests, not counted
Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.
- AA-OmniscienceClaude Opus 5 by 17.419.737.1+17.4
- GDPval-AA v2Claude Opus 5 by 142 rating points16821824+142 rating
- GDPval-AA v2 EloClaude Opus 5 by 138 rating points16861824+138 rating
- JobBenchClaude Opus 5 by 12.852.965.7+12.8
- Program BenchClaude Opus 5 by 7.677.885.4+7.6
- AA IntelligenceClaude Opus 5 by 7.243.650.8+7.2
- AA Agentic IndexClaude Opus 5 by 5.650.656.2+5.6
- livebench_data_analysisKimi K3 by 4.178.774.5+4.1
- Toolathlon VerifiedClaude Opus 5 by 4.176.580.6+4.1
- HLE (with tools)Claude Opus 5 by 3.859.863.6+3.8
- OfficeQA ProClaude Opus 5 by 3.663.366.9+3.6
- Agents' Last ExamClaude Opus 5 by 3.328.331.6+3.3
- livebench_languageClaude Opus 5 by 3.285.588.7+3.2
- ARC-AGI-1Claude Opus 5 by 394.597.5+3
- Artificial Analysis Coding IndexClaude Opus 5 by 1.876.278+1.8
- DeepSWE 1.1tie6968.8tie
Questions people ask
Which is better, Kimi K3 or Claude Opus 5?
Claude Opus 5 wins five of the eight areas we test: coding, agents, reasoning, facts and math. Kimi K3 wins long documents and following instructions, and costs 40% less. They are level on images and charts.
Which is better for coding?
Claude Opus 5. It wins 3 of the 5 coding tests both models report; Kimi K3 wins 1, and 1 is a tie.
Which is cheaper?
Kimi K3 costs $3.00 per million input tokens and $15.00 per million output tokens; Claude Opus 5 costs $5.00 and $25.00. That makes Kimi K3 about 40% cheaper for the same work.
How do you compare the two?
We use the 43 benchmark tests both models have published scores on. The verdict counts the 27 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 16 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.