Claude Sonnet 4.5 vs Kimi K2.6
Wins 0 of 8 areas
—
Wins 8 of 8 areas
Coding · Agents · Reasoning · Facts · Images and charts · Math · Long documents · Following instructions
Kimi K2.6 is the stronger all-rounder.
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.
- CodingWriting and fixing software08Kimi K2.68 of 8 tests
- AgentsCarrying out multi-step tasks on its own14Kimi K2.64 of 6 tests · 1 tie
- FactsGetting facts right instead of making them up03Kimi K2.63 of 4 tests · 1 tie
- ReasoningHard problems that need careful thinking03Kimi K2.63 of 3 tests
- MathCompetition and research-level math03Kimi K2.63 of 3 tests
- Images and chartsUnderstanding pictures, charts and video02Kimi K2.62 of 2 tests
- Long documentsFinding answers in very long texts01Kimi K2.61 of 1 test
- Following instructionsDoing exactly what it is asked01Kimi K2.61 of 1 test
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 K2.6 costs 73% less for the same work.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where Kimi K2.6 pulls ahead
- Finds hard-to-locate facts by browsing the webBrowseComp+62.2points ahead
- Unpublished advanced math problemsFrontierMath Tiers 1-3 (v2)+33.3points ahead
- Recent programming contest problemsLiveCodeBench v6+25.6points ahead
Where Claude Sonnet 4.5 pulls ahead
- Customer-service tasks in a simulated bankτ-Bench V3 · Banking+1.2points 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.
CodingKimi K2.6
- LiveCodeBench v6Kimi K2.6 by 25.66489.6+25.6
- SWE-bench ProKimi K2.6 by 1543.658.6+15
- LMArena · WebDevKimi K2.6 by 115 rating points13941509+115 rating
- Terminal-Bench 2.1Kimi K2.6 by 10.155.865.9+10.1
- SWE-bench MultilingualKimi K2.6 by 9.76776.7+9.7
- Terminal-Bench HardKimi K2.6 by 8.335.643.9+8.3
- SciCodeKimi K2.6 by 5.845.751.5+5.8
- SWE-bench VerifiedKimi K2.6 by 377.280.2+3
AgentsKimi K2.6
- BrowseCompKimi K2.6 by 62.224.186.3+62.2
- MCP AtlasKimi K2.6 by 24.343.868.1+24.3
- OSWorld-VerifiedKimi K2.6 by 11.761.473.1+11.7
- GDPValKimi K2.6 by 6.520.527+6.5
- τ-Bench V3 · BankingClaude Sonnet 4.5 by 1.224.523.3+1.2
- Terminal-Bench 4.0tie00.5tie
ReasoningKimi K2.6
- Humanity's Last ExamKimi K2.6 by 19.717.837.5+19.7
- GPQA DiamondKimi K2.6 by 7.783.491.1+7.7
- CritPtKimi K2.6 by 6.91.18+6.9
FactsKimi K2.6
- AA-Omniscience · Non-hallucinationKimi K2.6 by 8.650.959.5+8.6
- SimpleQA VerifiedKimi K2.6 by 4.230.734.9+4.2
- Vectara HHEM hallucination ratelower is betterKimi K2.6 by 1.21210.8+1.2
- AA-Omniscience · Accuracytie32.932.6tie
Images and chartsKimi K2.6
- CharXiv (RQ)Kimi K2.6 by 19.567.286.7+19.5
- MMMU-ProKimi K2.6 by 10.768.779.4+10.7
MathKimi K2.6
- FrontierMath Tiers 1-3 (v2)Kimi K2.6 by 33.323.957.2+33.3
- FrontierMath Tier 4Kimi K2.6 by 23.22.425.6+23.2
- IMOAnswerBenchKimi K2.6 by 20.165.986+20.1
Following instructionsKimi K2.6
- IFBenchKimi K2.6 by 18.757.376+18.7
Other results15 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.
- HLE (with tools)Kimi K2.6 by 223254+22
- LiveCodeBenchKimi K2.6 by 18.67189.6+18.6
- τ²-Bench Telecom (AA run)Kimi K2.6 by 17.878.195.9+17.8
- Terminal-Bench 2.0Kimi K2.6 by 16.75066.7+16.7
- frontiermath_tier_4_v1Kimi K2.6 by 10.44.214.6+10.4
- Artificial Analysis Coding IndexKimi K2.6 by 9.752.161.8+9.7
- ToolathlonKimi K2.6 by 94150+9
- GPQA (unspecified)Kimi K2.6 by 7.683.491+7.6
- AA IntelligenceKimi K2.6 by 6.320.727+6.3
- AA-OmniscienceKimi K2.6 by 5.4-0.15.3+5.4
- AA Agentic IndexKimi K2.6 by 4.617.522.1+4.6
- vectara_answer_rateKimi K2.6 by 4.195.699.7+4.1
- vectara_factual_consistencyKimi K2.6 by 1.28889.2+1.2
- vectara_avg_summary_lengthClaude Sonnet 4.5 by 11.1 rating points127.8116.7+11.1 rating
- MMLU-Protie87.587.1tie
Questions people ask
Which is better, Claude Sonnet 4.5 or Kimi K2.6?
Kimi K2.6 wins all eight areas we test: coding, agents, reasoning, facts, images and charts, math, long documents and following instructions. Claude Sonnet 4.5 wins none.
Which is better for coding?
Kimi K2.6. It wins 8 of the 8 coding tests both models report; Claude Sonnet 4.5 wins none.
Which is cheaper?
Claude Sonnet 4.5 costs $3.00 per million input tokens and $15.00 per million output tokens; Kimi K2.6 costs $0.95 and $4.00. That makes Kimi K2.6 about 73% 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 28 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 15 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.