DeepSeek-V3.2 vs Qwen3 Max (Reasoning)
Wins 1 of 7 areas
Facts
Wins 5 of 7 areas
Agents · Reasoning · Math · Long documents · Following instructions
Qwen3 Max (Reasoning) is the stronger all-rounder.DeepSeek-V3.2 is cheaper and better at facts.
Scores updated · 27 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.
- ReasoningHard problems that need careful thinking12Qwen3 Max (Reasoning)2 of 3 tests
- MathCompetition and research-level math01Qwen3 Max (Reasoning)1 of 2 tests · 1 tie
- Long documentsFinding answers in very long texts01Qwen3 Max (Reasoning)1 of 2 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own01Qwen3 Max (Reasoning)1 of 1 test
- Following instructionsDoing exactly what it is asked01Qwen3 Max (Reasoning)1 of 1 test
- FactsGetting facts right instead of making them up20DeepSeek-V3.22 of 2 tests
- CodingWriting and fixing software22Even2 each
Agents 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.
DeepSeek-V3.2 costs 90% 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 Max (Reasoning) pulls ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+10points ahead
- Olympiad math problems with short answersIMOAnswerBench+5.6points ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+3.4points ahead
Where DeepSeek-V3.2 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+12points ahead
- Hard command-line tasks in a real terminalTerminal-Bench Hard+11.4points ahead
- Fixes real GitHub issues in many programming languagesSWE-bench Multilingual+3.5points ahead
Every test, side by side
All 27 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingEven
- Terminal-Bench HardDeepSeek-V3.2 by 11.435.624.2+11.4
- SWE-bench MultilingualDeepSeek-V3.2 by 3.570.266.7+3.5
- LiveCodeBench v6Qwen3 Max (Reasoning) by 2.683.385.9+2.6
- SWE-bench VerifiedQwen3 Max (Reasoning) by 2.273.175.3+2.2
ReasoningQwen3 Max (Reasoning)
- Humanity's Last ExamQwen3 Max (Reasoning) by 3.424.628+3.4
- GPQA DiamondQwen3 Max (Reasoning) by 2.18486.1+2.1
- CritPtDeepSeek-V3.2 by 1.22.91.7+1.2
FactsDeepSeek-V3.2
- AA-Omniscience · Non-hallucinationDeepSeek-V3.2 by 1217.35.3+12
- AA-Omniscience · AccuracyDeepSeek-V3.2 by 3.23329.8+3.2
MathQwen3 Max (Reasoning)
- IMOAnswerBenchQwen3 Max (Reasoning) by 5.678.383.9+5.6
- AIME 2026tie94.293.3tie
Long documentsQwen3 Max (Reasoning)
- AA-LCRQwen3 Max (Reasoning) by 173.374.3+1
- LongBench v2tie59.860.6tie
Following instructionsQwen3 Max (Reasoning)
- IFBenchQwen3 Max (Reasoning) by 1060.770.7+10
Other results12 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.
- Terminal-Bench 2.0DeepSeek-V3.2 by 23.946.422.5+23.9
- ToolathlonDeepSeek-V3.2 by 16.435.218.8+16.4
- AA-OmniscienceDeepSeek-V3.2 by 14.1-22.5-36.6+14.1
- AIME 2025DeepSeek-V3.2 by 11.593.181.6+11.5
- τ²-Bench Telecom (AA run)DeepSeek-V3.2 by 790.683.6+7
- t2-benchDeepSeek-V3.2 by 5.480.274.8+5.4
- MCPMarkDeepSeek-V3.2 by 4.53833.5+4.5
- BrowseComp-ZHDeepSeek-V3.2 by 4.16560.9+4.1
- Seal-0DeepSeek-V3.2 by 2.649.546.9+2.6
- MMLU-Reduxtie93.792.8tie
- MMLU-Protie8585.7tie
- AA Intelligencetie21.521.3tie
Questions people ask
Which is better, DeepSeek-V3.2 or Qwen3 Max (Reasoning)?
Qwen3 Max (Reasoning) wins five of the seven areas where both have results: agents, reasoning, math, long documents and following instructions. DeepSeek-V3.2 wins facts, and costs 90% less. They are level on coding.
Which is better for coding?
Neither. They win 2 coding tests each of the 4 both models report.
Which is cheaper?
DeepSeek-V3.2 costs $0.28 per million input tokens and $0.42 per million output tokens; Qwen3 Max (Reasoning) costs $1.20 and $6.00. That makes DeepSeek-V3.2 about 90% cheaper for the same work.
How do you compare the two?
We use the 27 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 12 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.