Claude Sonnet 4 vs Qwen3.5 397B A17B
Wins 0 of 7 areas
—
Wins 6 of 7 areas
Coding · Agents · Reasoning · Images and charts · Long documents · Following instructions
Qwen3.5 397B A17B is the stronger all-rounder.
Scores updated · 34 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 software07Qwen3.5 397B A17B7 of 7 tests
- ReasoningHard problems that need careful thinking03Qwen3.5 397B A17B3 of 3 tests
- Images and chartsUnderstanding pictures, charts and video03Qwen3.5 397B A17B3 of 3 tests
- Following instructionsDoing exactly what it is asked02Qwen3.5 397B A17B2 of 2 tests
- AgentsCarrying out multi-step tasks on its own12Qwen3.5 397B A17B2 of 3 tests
- Long documentsFinding answers in very long texts01Qwen3.5 397B A17B1 of 1 test
- FactsGetting facts right instead of making them up11Even1 each
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 77% 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 Qwen3.5 397B A17B pulls ahead
- Finds hard-to-locate facts by browsing the webBrowseComp+56.8points ahead
- Recent programming contest problemsLiveCodeBench v6+35.1points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+24.1points ahead
Where Claude Sonnet 4 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+59.8points ahead
- Customer-service tasks in a simulated bankτ-Bench V3 · Banking+3.3points ahead
Every test, side by side
All 34 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 35.148.583.6+35.1
- Terminal-Bench 2.1Qwen3.5 397B A17B by 1536.351.3+15
- SWE-bench MultilingualQwen3.5 397B A17B by 12.456.969.3+12.4
- Terminal-Bench HardQwen3.5 397B A17B by 9.831.140.9+9.8
- SWE-bench ProQwen3.5 397B A17B by 8.242.750.9+8.2
- SciCodeQwen3.5 397B A17B by 4.84044.8+4.8
- SWE-bench VerifiedQwen3.5 397B A17B by 3.772.776.4+3.7
AgentsQwen3.5 397B A17B
- BrowseCompQwen3.5 397B A17B by 56.812.269+56.8
- GDPValQwen3.5 397B A17B by 59.614.6+5
- τ-Bench V3 · BankingClaude Sonnet 4 by 3.316.713.4+3.3
ReasoningQwen3.5 397B A17B
- Humanity's Last ExamQwen3.5 397B A17B by 18.310.729+18.3
- GPQA DiamondQwen3.5 397B A17B by 11.677.789.3+11.6
- CritPtQwen3.5 397B A17B by 1.40.31.7+1.4
FactsEven
- AA-Omniscience · Non-hallucinationClaude Sonnet 4 by 59.870.911.1+59.8
- AA-Omniscience · AccuracyQwen3.5 397B A17B by 8.122.730.8+8.1
Images and chartsQwen3.5 397B A17B
- MMMU-ProQwen3.5 397B A17B by 15.561.877.3+15.5
- MMMUQwen3.5 397B A17B by 10.674.485+10.6
- LMArena · VisionQwen3.5 397B A17B by 71 rating points11911262+71 rating
Following instructionsQwen3.5 397B A17B
- IFBenchQwen3.5 397B A17B by 24.154.778.8+24.1
- Multi-ChallengeQwen3.5 397B A17B by 10.557.167.6+10.5
Other results13 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.
- HMMT 2025Qwen3.5 397B A17B by 76.815.992.7+76.8
- BrowseComp-ZHQwen3.5 397B A17B by 41.229.170.3+41.2
- τ²-Bench Telecom (AA run)Qwen3.5 397B A17B by 3164.695.6+31
- AA-OmniscienceClaude Sonnet 4 by 30.90.2-30.8+30.9
- HLE (with tools)Qwen3.5 397B A17B by 2820.348.3+28
- SuperGPQAQwen3.5 397B A17B by 14.755.770.4+14.7
- Artificial Analysis Coding IndexQwen3.5 397B A17B by 10.637.648.2+10.6
- AA Agentic IndexClaude Sonnet 4 by 717.610.6+7
- IFEvalQwen3.5 397B A17B by 587.692.6+5
- MMLU-ProQwen3.5 397B A17B by 4.183.787.8+4.1
- MMMLUQwen3.5 397B A17B by 286.588.5+2
- MMLU-ReduxQwen3.5 397B A17B by 1.393.694.9+1.3
- AA Intelligencetie18.918.4tie
Questions people ask
Which is better, Claude Sonnet 4 or Qwen3.5 397B A17B?
Qwen3.5 397B A17B wins six of the seven areas where both have results: coding, agents, reasoning, images and charts, long documents and following instructions. Claude Sonnet 4 wins none. They are level on facts.
Which is better for coding?
Qwen3.5 397B A17B. It wins 7 of the 7 coding tests both models report; Claude Sonnet 4 wins none.
Which is cheaper?
Claude Sonnet 4 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 34 benchmark tests both models have published scores on. The verdict counts the 21 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 13 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.