Claude Sonnet 4 vs Qwen3 235B A22B
Wins 5 of 6 areas
Coding · Reasoning · Facts · Long documents · Following instructions
Wins 1 of 6 areas
Agents
Claude Sonnet 4 is the stronger all-rounder.Qwen3 235B A22B is cheaper and better at agents.
Scores updated · 51 tests both models report · How we compare
Where each one wins
Tests won in each of the six 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 software40Claude Sonnet 44 of 5 tests · 1 tie
- Following instructionsDoing exactly what it is asked20Claude Sonnet 42 of 2 tests
- ReasoningHard problems that need careful thinking10Claude Sonnet 41 of 3 tests · 2 ties
- FactsGetting facts right instead of making them up21Claude Sonnet 42 of 3 tests
- Long documentsFinding answers in very long texts10Claude Sonnet 41 of 1 test
- AgentsCarrying out multi-step tasks on its own01Qwen3 235B A22B1 of 1 test
Agents 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 235B A22B costs 81% 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 Sonnet 4 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+48.5points ahead
- Fixes real GitHub issues in Python projectsSWE-bench Verified+38.3points ahead
- Hard command-line tasks in a real terminalTerminal-Bench Hard+25points ahead
Where Qwen3 235B A22B pulls ahead
- Real work tasks from 44 professionsGDPVal+2.1points ahead
Every test, side by side
All 51 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Sonnet 4
- SWE-bench VerifiedClaude Sonnet 4 by 38.372.734.4+38.3
- Terminal-Bench HardClaude Sonnet 4 by 2531.16.1+25
- SWE-bench ProClaude Sonnet 4 by 21.342.721.4+21.3
- LiveCodeBench v6Claude Sonnet 4 by 11.548.537+11.5
- SciCodetie4039.9tie
ReasoningClaude Sonnet 4
- GPQA DiamondClaude Sonnet 4 by 7.777.770+7.7
- CritPttie0.30tie
- Humanity's Last Examtie10.711tie
FactsClaude Sonnet 4
- AA-Omniscience · Non-hallucinationClaude Sonnet 4 by 48.570.922.4+48.5
- AA-Omniscience · AccuracyClaude Sonnet 4 by 4.222.718.5+4.2
- Vectara HHEM hallucination ratelower is betterQwen3 235B A22B by 110.39.3+1
Following instructionsClaude Sonnet 4
- IFBenchClaude Sonnet 4 by 1654.738.7+16
- Multi-ChallengeClaude Sonnet 4 by 15.957.141.2+15.9
Other results36 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 Sonnet 4 by 44.90.2-44.7+44.9
- AIME 2024Qwen3 235B A22B by 42.343.485.7+42.3
- τ²-Bench Telecom (AA run)Claude Sonnet 4 by 40.664.624+40.6
- ZebraLogicClaude Sonnet 4 by 3673.737.7+36
- Tau2 airlineClaude Sonnet 4 by 2955.526.5+29
- TAU-bench (airline)Claude Sonnet 4 by 25.36034.7+25.3
- τ²-BenchClaude Sonnet 4 by 23.145.222.1+23.1
- MultiPL-EClaude Sonnet 4 by 22.788.665.9+22.7
- TAU-bench (retail)Claude Sonnet 4 by 21.980.558.6+21.9
- livecodebench_hardQwen3 235B A22B by 21.432.654+21.4
- Artificial Analysis Coding IndexClaude Sonnet 4 by 20.237.617.4+20.2
- τ²-Bench (Retail)Claude Sonnet 4 by 187557+18
- MMLU-ProClaude Sonnet 4 by 15.583.768.2+15.5
- livecodebench_mediumQwen3 235B A22B by 12.87688.8+12.8
- CNMO 2024Claude Sonnet 4 by 11.860.448.6+11.8
- SuperGPQAClaude Sonnet 4 by 11.655.744.1+11.6
- AIME 2025Qwen3 235B A22B by 1170.581.5+11
- AA IntelligenceClaude Sonnet 4 by 9.418.99.5+9.4
- AutoLogiClaude Sonnet 4 by 6.589.883.3+6.5
- MMLU-ReduxClaude Sonnet 4 by 6.293.687.4+6.2
- AceBenchClaude Sonnet 4 by 5.776.270.5+5.7
- MMLUClaude Sonnet 4 by 4.591.587+4.5
- IFEvalClaude Sonnet 4 by 4.487.683.2+4.4
- vectara_avg_summary_lengthClaude Sonnet 4 by 40.2 rating points145.8105.6+40.2 rating
- HMMT 2025Claude Sonnet 4 by 415.911.9+4
- OJBenchClaude Sonnet 4 by 415.311.3+4
- vectara_answer_rateClaude Sonnet 4 by 3.798.694.9+3.7
- MATH-500 (EM)Claude Sonnet 4 by 2.89491.2+2.8
- SimpleQAClaude Sonnet 4 by 2.715.913.2+2.7
- LiveBenchQwen3 235B A22B by 2.374.877.1+2.3
- LiveCodeBenchQwen3 235B A22B by 2.268.570.7+2.2
- vectara_factual_consistencyQwen3 235B A22B by 189.790.7+1
- PolyMath-entie52.851.9tie
- AA Agentic Indextie17.618.4tie
- livecodebench_easytie98.899.1tie
- MMMLUtie86.586.7tie
Questions people ask
Which is better, Claude Sonnet 4 or Qwen3 235B A22B?
Claude Sonnet 4 wins five of the six areas where both have results: coding, reasoning, facts, long documents and following instructions. Qwen3 235B A22B wins agents, and costs 81% less.
Which is better for coding?
Claude Sonnet 4. It wins 4 of the 5 coding tests both models report; Qwen3 235B A22B wins none, and 1 is a tie.
Which is cheaper?
Claude Sonnet 4 costs $3.00 per million input tokens and $15.00 per million output tokens; Qwen3 235B A22B costs $0.70 and $2.80. That makes Qwen3 235B A22B about 81% cheaper for the same work.
How do you compare the two?
We use the 51 benchmark tests both models have published scores on. The verdict counts the 15 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 36 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.