Claude Opus 4.8 vs Kimi K2.7 Code
Wins 6 of 7 areas
Coding · Agents · Reasoning · Facts · Math · Following instructions
Wins 1 of 7 areas
Long documents
Claude Opus 4.8 is the stronger all-rounder.Kimi K2.7 Code is cheaper and better at long documents.
Scores updated · 36 tests both models report · How we compare
Where each one wins
Tests won in each of the seven areas where both have results. 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 software60Claude Opus 4.86 of 6 tests
- ReasoningHard problems that need careful thinking50Claude Opus 4.85 of 5 tests
- AgentsCarrying out multi-step tasks on its own40Claude Opus 4.84 of 4 tests
- FactsGetting facts right instead of making them up30Claude Opus 4.83 of 3 tests
- MathCompetition and research-level math30Claude Opus 4.83 of 3 tests
- Following instructionsDoing exactly what it is asked10Claude Opus 4.81 of 2 tests · 1 tie
- Long documentsFinding answers in very long texts01Kimi K2.7 Code1 of 1 test
Long documents rests on a single test.
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 K2.7 Code costs 84% 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 Claude Opus 4.8 pulls ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+43.9points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+43.1points ahead
- Unpublished advanced math problemsFrontierMath Tiers 1-3 (v2)+26points ahead
Where Kimi K2.7 Code pulls ahead
- Reasons across sets of long documentsAA-LCR+1.6points ahead
Every test, side by side
All 36 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Opus 4.8
- Terminal-Bench 2.1Claude Opus 4.8 by 17.284.667.4+17.2
- Terminal-Bench HardClaude Opus 4.8 by 13.658.344.7+13.6
- LMArena · WebDevClaude Opus 4.8 by 83 rating points15561473+83 rating
- LiveBench · CodingClaude Opus 4.8 by 7.881.874+7.8
- SciCodeClaude Opus 4.8 by 6.654.447.8+6.6
- LiveBench · Agentic CodingClaude Opus 4.8 by 4.850.545.7+4.8
AgentsClaude Opus 4.8
- GDPValClaude Opus 4.8 by 20.847.827+20.8
- Terminal-Bench 4.0Claude Opus 4.8 by 20.721.71+20.7
- τ-Bench V3 · BankingClaude Opus 4.8 by 1434.220.2+14
- MCP AtlasClaude Opus 4.8 by 6.282.276+6.2
ReasoningClaude Opus 4.8
- Humanity's Last ExamClaude Opus 4.8 by 13.748.735+13.7
- CritPtClaude Opus 4.8 by 10.920.910+10.9
- SimpleBenchClaude Opus 4.8 by 6.964.857.9+6.9
- LiveBench · ReasoningClaude Opus 4.8 by 6.489.282.8+6.4
- GPQA DiamondClaude Opus 4.8 by 2.49289.6+2.4
FactsClaude Opus 4.8
- AA-Omniscience · Non-hallucinationClaude Opus 4.8 by 43.160.717.6+43.1
- SimpleQA VerifiedClaude Opus 4.8 by 16.55336.5+16.5
- AA-Omniscience · AccuracyClaude Opus 4.8 by 9.248.839.6+9.2
MathClaude Opus 4.8
- FrontierMath Tier 4Claude Opus 4.8 by 43.956.112.2+43.9
- FrontierMath Tiers 1-3 (v2)Claude Opus 4.8 by 268054+26
- LiveBench · MathematicsClaude Opus 4.8 by 14.794.379.6+14.7
Following instructionsClaude Opus 4.8
- LiveBench · Instruction FollowingClaude Opus 4.8 by 15.77256.3+15.7
- IFBenchtie62.263.1tie
Other results12 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 4.8 by 3928.8-10.2+39
- DeepSWE 1.1Claude Opus 4.8 by 285931+28
- AA Agentic IndexClaude Opus 4.8 by 20.142.622.5+20.1
- Program BenchClaude Opus 4.8 by 18.371.953.6+18.3
- AA IntelligenceClaude Opus 4.8 by 1641.825.8+16
- Artificial Analysis Coding IndexClaude Opus 4.8 by 13.574.360.8+13.5
- MLS Bench Litelower is betterKimi K2.7 Code by 7.742.835.1+7.7
- LiveBenchClaude Opus 4.8 by 5.377.271.9+5.3
- MCP Mark VerifiedKimi K2.7 Code by 4.776.481.1+4.7
- τ²-Bench Telecom (AA run)Claude Opus 4.8 by 4.394.490.1+4.3
- livebench_data_analysisClaude Opus 4.8 by 3.36662.7+3.3
- livebench_languageClaude Opus 4.8 by 1.879.777.9+1.8
Questions people ask
Which is better, Claude Opus 4.8 or Kimi K2.7 Code?
Claude Opus 4.8 wins six of the seven areas where both have results: coding, agents, reasoning, facts, math and following instructions. Kimi K2.7 Code wins long documents, and costs 84% less.
Which is better for coding?
Claude Opus 4.8. It wins 6 of the 6 coding tests both models report; Kimi K2.7 Code wins none.
Which is cheaper?
Claude Opus 4.8 costs $5.00 per million input tokens and $25.00 per million output tokens; Kimi K2.7 Code costs $0.95 and $4.00. That makes Kimi K2.7 Code about 84% cheaper for the same work.
How do you compare the two?
We use the 36 benchmark tests both models have published scores on. The verdict counts the 24 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 12 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.