Claude Sonnet 4.6 vs Qwen3.6 Plus
Wins 5 of 8 areas
Coding · Agents · Reasoning · Math · Long documents
Wins 2 of 8 areas
Facts · Images and charts
Claude Sonnet 4.6 is the stronger all-rounder.Qwen3.6 Plus is cheaper and better at images and charts.
Scores updated · 47 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 Sonnet 4.67 of 8 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own31Claude Sonnet 4.63 of 4 tests
- ReasoningHard problems that need careful thinking20Claude Sonnet 4.62 of 4 tests · 2 ties
- MathCompetition and research-level math10Claude Sonnet 4.61 of 1 test
- Long documentsFinding answers in very long texts10Claude Sonnet 4.61 of 1 test
- Images and chartsUnderstanding pictures, charts and video02Qwen3.6 Plus2 of 2 tests
- FactsGetting facts right instead of making them up12Qwen3.6 Plus2 of 3 tests
- Following instructionsDoing exactly what it is asked11Even1 each
Math and long documents 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.
Qwen3.6 Plus costs 81% 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 Sonnet 4.6 pulls ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+14.5points ahead
- Customer-service tasks in a simulated bankτ-Bench V3 · Banking+13.6points ahead
- Real work tasks from 44 professionsGDPVal+12points ahead
Where Qwen3.6 Plus pulls ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+18.6points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+13.8points ahead
- Multi-step tasks using many real software toolsMCP Atlas+12.8points ahead
Every test, side by side
All 47 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Sonnet 4.6
- Terminal-Bench 2.1Claude Sonnet 4.6 by 9.871.261.4+9.8
- SciCodeClaude Sonnet 4.6 by 9.450.140.7+9.4
- Terminal-Bench HardClaude Sonnet 4.6 by 9.15343.9+9.1
- LMArena · WebDevClaude Sonnet 4.6 by 61 rating points15221461+61 rating
- SWE-bench ProClaude Sonnet 4.6 by 1.558.156.6+1.5
- LiveBench · Agentic CodingClaude Sonnet 4.6 by 1.242.641.4+1.2
- LiveBench · CodingClaude Sonnet 4.6 by 1.179.378.2+1.1
- SWE-bench Verifiedtie79.678.8tie
AgentsClaude Sonnet 4.6
- τ-Bench V3 · BankingClaude Sonnet 4.6 by 13.634.420.8+13.6
- MCP AtlasQwen3.6 Plus by 12.861.374.1+12.8
- GDPValClaude Sonnet 4.6 by 1236.724.7+12
- OSWorld-VerifiedClaude Sonnet 4.6 by 1072.562.5+10
ReasoningClaude Sonnet 4.6
- LiveBench · ReasoningClaude Sonnet 4.6 by 984.875.8+9
- Humanity's Last ExamClaude Sonnet 4.6 by 5.833.627.8+5.8
- GPQA Diamondtie87.588.2tie
- CritPttie3.12.9tie
FactsQwen3.6 Plus
- AA-Omniscience · AccuracyClaude Sonnet 4.6 by 14.540.926.4+14.5
- AA-Omniscience · Non-hallucinationQwen3.6 Plus by 13.851.665.4+13.8
- SimpleQA VerifiedQwen3.6 Plus by 11.332.844.1+11.3
Images and chartsQwen3.6 Plus
- CharXiv (RQ)Qwen3.6 Plus by 9.971.681.5+9.9
- MMMU-ProQwen3.6 Plus by 4.773.378+4.7
Following instructionsEven
- IFBenchQwen3.6 Plus by 18.656.675.2+18.6
- LiveBench · Instruction FollowingClaude Sonnet 4.6 by 4.963.258.3+4.9
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.
- τ²-Bench Telecom (AA run)Qwen3.6 Plus by 2275.797.7+22
- AgentWorldBench - MCPClaude Sonnet 4.6 by 14.77055.3+14.7
- AA-OmniscienceClaude Sonnet 4.6 by 11.312.20.9+11.3
- Finance Agent v2Claude Sonnet 4.6 by 10.15140.9+10.1
- Artificial Analysis Coding IndexClaude Sonnet 4.6 by 8.56354.5+8.5
- livebench_data_analysisClaude Sonnet 4.6 by 8.17869.9+8.1
- AgentWorldBench - SearchClaude Sonnet 4.6 by 6.928.821.9+6.9
- AgentWorldBench - TerminalClaude Sonnet 4.6 by 6.45750.6+6.4
- AgentWorldBench - SWEClaude Sonnet 4.6 by 5.464.559.1+5.4
- AgentWorldBench - OverallClaude Sonnet 4.6 by 5.25650.8+5.2
- LiveBenchClaude Sonnet 4.6 by 4.675.570.8+4.6
- AA Agentic IndexClaude Sonnet 4.6 by 4.133.129+4.1
- HLE (with tools)Qwen3.6 Plus by 3.846.850.6+3.8
- AA IntelligenceClaude Sonnet 4.6 by 3.130.127+3.1
- AgentWorldBench - OSClaude Sonnet 4.6 by 2.963.260.3+2.9
- Terminal-Bench 2.0Qwen3.6 Plus by 2.559.161.6+2.5
- ToolathlonClaude Sonnet 4.6 by 1.24139.8+1.2
- livebench_languageClaude Sonnet 4.6 by 1.176.175+1.1
- AgentWorldBench - Androidtie5857.6tie
- MMMLUtie89.389.5tie
- AgentWorldBench - Webtie50.850.8tie
- frontiermath_tier_4_v1tie8.38.3tie
Questions people ask
Which is better, Claude Sonnet 4.6 or Qwen3.6 Plus?
Claude Sonnet 4.6 wins five of the eight areas we test: coding, agents, reasoning, math and long documents. Qwen3.6 Plus wins facts and images and charts, and costs 81% less. They are level on following instructions.
Which is better for coding?
Claude Sonnet 4.6. It wins 7 of the 8 coding tests both models report; Qwen3.6 Plus wins none, and 1 is a tie.
Which is cheaper?
Claude Sonnet 4.6 costs $3.00 per million input tokens and $15.00 per million output tokens; Qwen3.6 Plus costs $0.50 and $3.00. That makes Qwen3.6 Plus about 81% cheaper for the same work.
How do you compare the two?
We use the 47 benchmark tests both models have published scores on. The verdict counts the 25 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.