Claude Sonnet 4.5 vs Qwen3.5 397B A17B
Wins 2 of 8 areas
Agents · Facts
Wins 6 of 8 areas
Coding · Reasoning · Images and charts · Math · Long documents · Following instructions
Qwen3.5 397B A17B is the stronger all-rounder.Claude Sonnet 4.5 is better at facts.
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 software14Qwen3.5 397B A17B4 of 8 tests · 3 ties
- ReasoningHard problems that need careful thinking02Qwen3.5 397B A17B2 of 3 tests · 1 tie
- Images and chartsUnderstanding pictures, charts and video02Qwen3.5 397B A17B2 of 2 tests
- MathCompetition and research-level math02Qwen3.5 397B A17B2 of 2 tests
- Long documentsFinding answers in very long texts02Qwen3.5 397B A17B2 of 2 tests
- Following instructionsDoing exactly what it is asked02Qwen3.5 397B A17B2 of 2 tests
- FactsGetting facts right instead of making them up30Claude Sonnet 4.53 of 3 tests
- AgentsCarrying out multi-step tasks on its own21Claude Sonnet 4.52 of 4 tests · 1 tie
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 77% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where Qwen3.5 397B A17B pulls ahead
- Finds hard-to-locate facts by browsing the webBrowseComp+44.9points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+21.5points ahead
- Recent programming contest problemsLiveCodeBench v6+19.6points ahead
Where Claude Sonnet 4.5 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+39.8points ahead
- Customer-service tasks in a simulated bankτ-Bench V3 · Banking+11.1points ahead
- Real work tasks from 44 professionsGDPVal+5.7points 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.
CodingQwen3.5 397B A17B
- LiveCodeBench v6Qwen3.5 397B A17B by 19.66483.6+19.6
- SWE-bench ProQwen3.5 397B A17B by 7.343.650.9+7.3
- Terminal-Bench HardQwen3.5 397B A17B by 5.335.640.9+5.3
- Terminal-Bench 2.1Claude Sonnet 4.5 by 4.555.851.3+4.5
- SWE-bench MultilingualQwen3.5 397B A17B by 2.36769.3+2.3
- SciCodetie45.744.8tie
- SWE-bench Verifiedtie77.276.4tie
- LMArena · WebDevtie13931399tie
AgentsClaude Sonnet 4.5
- BrowseCompQwen3.5 397B A17B by 44.924.169+44.9
- τ-Bench V3 · BankingClaude Sonnet 4.5 by 11.124.513.4+11.1
- GDPValClaude Sonnet 4.5 by 5.720.514.8+5.7
- Terminal-Bench 4.0tie00tie
ReasoningQwen3.5 397B A17B
- Humanity's Last ExamQwen3.5 397B A17B by 11.217.829+11.2
- GPQA DiamondQwen3.5 397B A17B by 5.983.489.3+5.9
- CritPttie1.11.7tie
FactsClaude Sonnet 4.5
- AA-Omniscience · Non-hallucinationClaude Sonnet 4.5 by 39.850.911.1+39.8
- SimpleQA VerifiedClaude Sonnet 4.5 by 4.730.726+4.7
- AA-Omniscience · AccuracyClaude Sonnet 4.5 by 2.132.930.8+2.1
Images and chartsQwen3.5 397B A17B
MathQwen3.5 397B A17B
- IMOAnswerBenchQwen3.5 397B A17B by 1565.980.9+15
- FrontierMath Tiers 1-3 (v2)Qwen3.5 397B A17B by 7.323.931.2+7.3
Long documentsQwen3.5 397B A17B
- AA-LCRQwen3.5 397B A17B by 572.377.3+5
- LongBench v2Qwen3.5 397B A17B by 1.461.863.2+1.4
Following instructionsQwen3.5 397B A17B
- IFBenchQwen3.5 397B A17B by 21.557.378.8+21.5
- Multi-ChallengeQwen3.5 397B A17B by 12.355.367.6+12.3
Other results17 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.
- BrowseComp (context management)Qwen3.5 397B A17B by 52.526.178.6+52.5
- AA-OmniscienceClaude Sonnet 4.5 by 30.6-0.1-30.8+30.6
- BrowseComp-ZHQwen3.5 397B A17B by 27.942.470.3+27.9
- HMMT Feb. 2025Qwen3.5 397B A17B by 27.367.594.8+27.3
- HMMT 2025Qwen3.5 397B A17B by 18.174.692.7+18.1
- τ²-Bench Telecom (AA run)Qwen3.5 397B A17B by 17.578.195.6+17.5
- HLE (with tools)Qwen3.5 397B A17B by 16.33248.3+16.3
- HMMT Nov. 2025Qwen3.5 397B A17B by 1181.792.7+11
- AA Agentic IndexClaude Sonnet 4.5 by 6.917.510.6+6.9
- Artificial Analysis Coding IndexClaude Sonnet 4.5 by 3.952.148.2+3.9
- GPQA (unspecified)Qwen3.5 397B A17B by 3.783.487.1+3.7
- ToolathlonClaude Sonnet 4.5 by 2.74138.3+2.7
- Terminal-Bench 2.0Qwen3.5 397B A17B by 2.55052.5+2.5
- AA IntelligenceClaude Sonnet 4.5 by 2.320.718.4+2.3
- MMLU-Reduxtie95.694.9tie
- MMMLUtie89.188.5tie
- MMLU-Protie87.587.8tie
Questions people ask
Which is better, Claude Sonnet 4.5 or Qwen3.5 397B A17B?
Qwen3.5 397B A17B wins six of the eight areas we test: coding, reasoning, images and charts, math, long documents and following instructions. Claude Sonnet 4.5 wins agents and facts.
Which is better for coding?
Qwen3.5 397B A17B. It wins 4 of the 8 coding tests both models report; Claude Sonnet 4.5 wins 1, and 3 are ties.
Which is cheaper?
Claude Sonnet 4.5 costs $3.00 per million input tokens and $15.00 per million output tokens; Qwen3.5 397B A17B costs $0.60 and $3.60. That makes Qwen3.5 397B A17B about 77% 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 26 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 17 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.