Claude Sonnet 4.5 vs Qwen3 30B A3B 2507 Thinking
Wins 6 of 6 areas
Coding · Agents · Reasoning · Facts · Long documents · Following instructions
Wins 0 of 6 areas
—
Claude Sonnet 4.5 is the stronger all-rounder.Qwen3 30B A3B 2507 Thinking is cheaper.
Scores updated · 24 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.
- AgentsCarrying out multi-step tasks on its own30Claude Sonnet 4.53 of 4 tests · 1 tie
- CodingWriting and fixing software31Claude Sonnet 4.53 of 4 tests
- ReasoningHard problems that need careful thinking20Claude Sonnet 4.52 of 3 tests · 1 tie
- FactsGetting facts right instead of making them up20Claude Sonnet 4.52 of 2 tests
- Long documentsFinding answers in very long texts10Claude Sonnet 4.51 of 1 test
- Following instructionsDoing exactly what it is asked10Claude Sonnet 4.51 of 1 test
Long documents and following instructions 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 30B A3B 2507 Thinking costs 86% 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 4.5 pulls ahead
- Fixes real GitHub issues in Python projectsSWE-bench Verified+55.2points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+37.4points ahead
- Hard command-line tasks in a real terminalTerminal-Bench Hard+30.3points ahead
Where Qwen3 30B A3B 2507 Thinking pulls ahead
- Recent programming contest problemsLiveCodeBench v6+2points ahead
Every test, side by side
All 24 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Sonnet 4.5
- SWE-bench VerifiedClaude Sonnet 4.5 by 55.277.222+55.2
- Terminal-Bench HardClaude Sonnet 4.5 by 30.335.65.3+30.3
- SciCodeClaude Sonnet 4.5 by 12.745.733+12.7
- LiveCodeBench v6Qwen3 30B A3B 2507 Thinking by 26466+2
AgentsClaude Sonnet 4.5
- BrowseCompClaude Sonnet 4.5 by 21.824.12.3+21.8
- GDPValClaude Sonnet 4.5 by 20.520.50+20.5
- τ-Bench V3 · BankingClaude Sonnet 4.5 by 19.124.55.4+19.1
- Terminal-Bench 4.0tie00tie
ReasoningClaude Sonnet 4.5
- GPQA DiamondClaude Sonnet 4.5 by 12.783.470.7+12.7
- Humanity's Last ExamClaude Sonnet 4.5 by 7.517.810.3+7.5
- CritPttie1.10.3tie
FactsClaude Sonnet 4.5
- AA-Omniscience · Non-hallucinationClaude Sonnet 4.5 by 37.450.913.5+37.4
- AA-Omniscience · AccuracyClaude Sonnet 4.5 by 16.632.916.3+16.6
Following instructionsClaude Sonnet 4.5
- IFBenchClaude Sonnet 4.5 by 6.657.350.7+6.6
Other results9 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 4.5 by 56-0.1-56.1+56
- τ²-Bench Telecom (AA run)Claude Sonnet 4.5 by 5078.128.1+50
- τ²-BenchClaude Sonnet 4.5 by 499849+49
- τ²-Bench (Retail)Claude Sonnet 4.5 by 30.186.256.1+30.1
- AIME25 no toolsClaude Sonnet 4.5 by 16.38871.7+16.3
- AA IntelligenceClaude Sonnet 4.5 by 10.920.79.8+10.9
- GPQA (unspecified)Claude Sonnet 4.5 by 1083.473.4+10
- MMLU-ProClaude Sonnet 4.5 by 5.687.581.9+5.6
- AIME 2025Claude Sonnet 4.5 by 28785+2
Questions people ask
Which is better, Claude Sonnet 4.5 or Qwen3 30B A3B 2507 Thinking?
Claude Sonnet 4.5 wins all six areas where both have results: coding, agents, reasoning, facts, long documents and following instructions. Qwen3 30B A3B 2507 Thinking wins none, but costs 86% less.
Which is better for coding?
Claude Sonnet 4.5. It wins 3 of the 4 coding tests both models report; Qwen3 30B A3B 2507 Thinking wins 1.
Which is cheaper?
Claude Sonnet 4.5 costs $3.00 per million input tokens and $15.00 per million output tokens; Qwen3 30B A3B 2507 Thinking costs $0.20 and $2.40. That makes Qwen3 30B A3B 2507 Thinking about 86% cheaper for the same work.
How do you compare the two?
We use the 24 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 9 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.