DeepSeek-V3 vs Qwen3.7 Max
Wins 0 of 6 areas
—
Wins 6 of 6 areas
Coding · Agents · Reasoning · Facts · Long documents · Following instructions
Qwen3.7 Max is the stronger all-rounder.DeepSeek-V3 is cheaper.
Scores updated · 27 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 software05Qwen3.7 Max5 of 5 tests
- ReasoningHard problems that need careful thinking04Qwen3.7 Max4 of 4 tests
- AgentsCarrying out multi-step tasks on its own03Qwen3.7 Max3 of 3 tests
- FactsGetting facts right instead of making them up02Qwen3.7 Max2 of 2 tests
- Long documentsFinding answers in very long texts01Qwen3.7 Max1 of 1 test
- Following instructionsDoing exactly what it is asked01Qwen3.7 Max1 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.
DeepSeek-V3 costs 89% 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 Qwen3.7 Max pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+60.3points ahead
- Command-line tasks in a real terminalTerminal-Bench 2.1+57.6points ahead
- Common-sense trick questionsSimpleBench+51.5points ahead
Where DeepSeek-V3 pulls ahead
No clear win on a test scored out of 100.
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.
CodingQwen3.7 Max
- Terminal-Bench 2.1Qwen3.7 Max by 57.616.974.5+57.6
- LiveCodeBench v6Qwen3.7 Max by 44.746.991.6+44.7
- SWE-bench VerifiedQwen3.7 Max by 38.44280.4+38.4
- Terminal-Bench HardQwen3.7 Max by 35.615.250.8+35.6
- SciCodeQwen3.7 Max by 10.53949.5+10.5
AgentsQwen3.7 Max
- GDPValQwen3.7 Max by 31.6031.6+31.6
- τ-Bench V3 · BankingQwen3.7 Max by 7.14.711.8+7.1
- Terminal-Bench 4.0Qwen3.7 Max by 1.501.5+1.5
ReasoningQwen3.7 Max
- SimpleBenchQwen3.7 Max by 51.518.970.4+51.5
- Humanity's Last ExamQwen3.7 Max by 35.84.740.5+35.8
- GPQA DiamondQwen3.7 Max by 26.865.592.3+26.8
- CritPtQwen3.7 Max by 13.4013.4+13.4
FactsQwen3.7 Max
- AA-Omniscience · Non-hallucinationQwen3.7 Max by 60.314.174.4+60.3
- AA-Omniscience · AccuracyQwen3.7 Max by 5.625.431.1+5.6
Following instructionsQwen3.7 Max
- IFBenchQwen3.7 Max by 39.54180.5+39.5
Other results11 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-OmniscienceQwen3.7 Max by 54.2-40.713.5+54.2
- τ²-Bench Telecom (AA run)Qwen3.7 Max by 47.647.194.7+47.6
- Artificial Analysis Coding IndexQwen3.7 Max by 432366+43
- SuperGPQAQwen3.7 Max by 19.953.773.6+19.9
- AA IntelligenceQwen3.7 Max by 19.89.729.5+19.8
- MMLU-ProXQwen3.7 Max by 16.570.587+16.5
- MMLU-ProQwen3.7 Max by 13.775.989.6+13.7
- MMMLUQwen3.7 Max by 10.979.490.3+10.9
- IFEvalQwen3.7 Max by 8.286.194.3+8.2
- MMLU-ReduxQwen3.7 Max by 5.989.195+5.9
- LiveBenchQwen3.7 Max by 1.972.474.3+1.9
Questions people ask
Which is better, DeepSeek-V3 or Qwen3.7 Max?
Qwen3.7 Max wins all six areas where both have results: coding, agents, reasoning, facts, long documents and following instructions. DeepSeek-V3 wins none, but costs 89% less.
Which is better for coding?
Qwen3.7 Max. It wins 5 of the 5 coding tests both models report; DeepSeek-V3 wins none.
Which is cheaper?
DeepSeek-V3 costs $0.24 per million input tokens and $0.90 per million output tokens; Qwen3.7 Max costs $2.50 and $7.50. That makes DeepSeek-V3 about 89% 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 16 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 11 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.