Claude Opus 4.6 vs Qwen3.5 397B A17B
Wins 5 of 8 areas
Coding · Agents · Reasoning · Facts · Math
Wins 1 of 8 areas
Following instructions
Claude Opus 4.6 is the stronger all-rounder.Qwen3.5 397B A17B is cheaper and better at following instructions.
Scores updated · 48 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 software70Claude Opus 4.67 of 7 tests
- MathCompetition and research-level math41Claude Opus 4.64 of 5 tests
- AgentsCarrying out multi-step tasks on its own30Claude Opus 4.63 of 3 tests
- FactsGetting facts right instead of making them up30Claude Opus 4.63 of 3 tests
- ReasoningHard problems that need careful thinking20Claude Opus 4.62 of 3 tests · 1 tie
- Following instructionsDoing exactly what it is asked02Qwen3.5 397B A17B2 of 2 tests
- Images and chartsUnderstanding pictures, charts and video11Even1 each
- Long documentsFinding answers in very long texts00Even0 each · 1 tie
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.
Qwen3.5 397B A17B costs 86% 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 4.6 pulls ahead
- Real work tasks from 44 professionsGDPVal+41.1points ahead
- Unpublished advanced math problemsFrontierMath Tiers 1-3 (v2)+34.8points ahead
- Proof problems from the US Math OlympiadUSAMO 2026+29.9points ahead
Where Qwen3.5 397B A17B pulls ahead
- Keeps track of context across a multi-turn chatMulti-Challenge+30.4points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+25.7points ahead
- Olympiad math problems with short answersIMOAnswerBench+5.6points ahead
Every test, side by side
All 48 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Opus 4.6
- LMArena · WebDevClaude Opus 4.6 by 146 rating points15461400+146 rating
- SWE-bench MultilingualClaude Opus 4.6 by 8.577.869.3+8.5
- SciCodeClaude Opus 4.6 by 7.151.944.8+7.1
- Terminal-Bench HardClaude Opus 4.6 by 5.346.240.9+5.3
- LiveCodeBench v6Claude Opus 4.6 by 5.288.883.6+5.2
- SWE-bench VerifiedClaude Opus 4.6 by 4.480.876.4+4.4
- SWE-bench ProClaude Opus 4.6 by 2.553.450.9+2.5
AgentsClaude Opus 4.6
- GDPValClaude Opus 4.6 by 41.155.914.8+41.1
- AA ApexAgentsClaude Opus 4.6 by 17.73315.3+17.7
- BrowseCompClaude Opus 4.6 by 158469+15
ReasoningClaude Opus 4.6
- CritPtClaude Opus 4.6 by 10.912.61.7+10.9
- Humanity's Last ExamClaude Opus 4.6 by 10.939.929+10.9
- GPQA Diamondtie89.689.3tie
FactsClaude Opus 4.6
- AA-Omniscience · Non-hallucinationClaude Opus 4.6 by 26.137.211.1+26.1
- SimpleQA VerifiedClaude Opus 4.6 by 214726+21
- AA-Omniscience · AccuracyClaude Opus 4.6 by 16.24730.8+16.2
Images and chartsEven
- LMArena · VisionClaude Opus 4.6 by 52 rating points13161264+52 rating
- MMMU-ProQwen3.5 397B A17B by 1.975.477.3+1.9
MathClaude Opus 4.6
- FrontierMath Tiers 1-3 (v2)Claude Opus 4.6 by 34.86631.2+34.8
- USAMO 2026Claude Opus 4.6 by 29.966.236.3+29.9
- HMMT Feb. 2026Claude Opus 4.6 by 8.396.287.9+8.3
- IMOAnswerBenchQwen3.5 397B A17B by 5.675.380.9+5.6
- AIME 2026Claude Opus 4.6 by 5.496.791.3+5.4
Following instructionsQwen3.5 397B A17B
- Multi-ChallengeQwen3.5 397B A17B by 30.437.167.6+30.4
- IFBenchQwen3.5 397B A17B by 25.753.178.8+25.7
Other results22 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 Agentic IndexClaude Opus 4.6 by 5767.610.6+57
- AA-OmniscienceClaude Opus 4.6 by 44.413.7-30.8+44.4
- ToolathlonClaude Opus 4.6 by 18.556.838.3+18.5
- NL2RepoClaude Opus 4.6 by 17.649.832.2+17.6
- FLTEvalClaude Opus 4.6 by 14.239.625.4+14.2
- AA IntelligenceClaude Opus 4.6 by 13.531.918.4+13.5
- Terminal-Bench 2.0Claude Opus 4.6 by 12.965.452.5+12.9
- MCPMarkClaude Opus 4.6 by 10.656.746.1+10.6
- AgentWorldBench - OSClaude Opus 4.6 by 9.370.260.9+9.3
- AgentWorldBench - AndroidClaude Opus 4.6 by 6.861.754.9+6.8
- HLE (with tools)Claude Opus 4.6 by 553.348.3+5
- HMMT Nov. 2025Claude Opus 4.6 by 3.696.392.7+3.6
- τ²-Bench Telecom (AA run)Qwen3.5 397B A17B by 3.592.195.6+3.5
- AgentWorldBench - OverallClaude Opus 4.6 by 3.157.854.7+3.1
- AgentWorldBench - WebClaude Opus 4.6 by 2.851.448.5+2.8
- MMMLUClaude Opus 4.6 by 2.691.188.5+2.6
- AgentWorldBench - TerminalClaude Opus 4.6 by 2.257.555.3+2.2
- AgentWorldBench - MCPClaude Opus 4.6 by 1.669.968.3+1.6
- AgentWorldBench - SearchQwen3.5 397B A17B by 1.529.330.8+1.5
- MMLU-ProClaude Opus 4.6 by 1.389.187.8+1.3
- AgentWorldBench - SWEtie64.564.4tie
- Artificial Analysis Coding Indextie48.148.2tie
Questions people ask
Which is better, Claude Opus 4.6 or Qwen3.5 397B A17B?
Claude Opus 4.6 wins five of the eight areas we test: coding, agents, reasoning, facts and math. Qwen3.5 397B A17B wins following instructions, and costs 86% less. They are level on images and charts and long documents.
Which is better for coding?
Claude Opus 4.6. It wins 7 of the 7 coding tests both models report; Qwen3.5 397B A17B wins none.
Which is cheaper?
Claude Opus 4.6 costs $5.00 per million input tokens and $25.00 per million output tokens; Qwen3.5 397B A17B costs $0.60 and $3.60. That makes Qwen3.5 397B A17B about 86% cheaper for the same work.
How do you compare the two?
We use the 48 benchmark tests both models have published scores on. The verdict counts the 26 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 22 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.