DeepSeek-V3 is the stronger all-rounder.
Scores updated · 26 tests both models report · How we compare
Where each one wins
Tests won in each of the two 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.
Coding 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.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where DeepSeek-V3 pulls ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+28.4points ahead
- Code for real scientific research problemsSciCode+16.1points ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+1points ahead
Where Qwen2 Instruct (72B) pulls ahead
No clear win on a test scored out of 100.
Every test, side by side
All 26 tests both models report. The winning score is in its model's colour; marks a score checked independently.
ReasoningDeepSeek-V3
- GPQA DiamondDeepSeek-V3 by 28.465.537.1+28.4
- Humanity's Last ExamDeepSeek-V3 by 14.73.7+1
Other results23 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.
- HarmBenchQwen2 Instruct (72B) by 27.149.776.8+27.1
- ARC-ChallengeDeepSeek-V3 by 26.495.368.9+26.4
- AIR-Bench 2024Qwen2 Instruct (72B) by 21.340.862.1+21.3
- HumanEvalQwen2 Instruct (72B) by 20.865.286+20.8
- MultiPL-EDeepSeek-V3 by 13.983.169.2+13.9
- MMLUDeepSeek-V3 by 11.688.576.9+11.6
- MMLU-ProDeepSeek-V3 by 11.575.964.4+11.5
- MATHDeepSeek-V3 by 11.290.279+11.2
- naturalquestions_closedbookDeepSeek-V3 by 7.746.739+7.7
- NarrativeQADeepSeek-V3 by 6.979.672.7+6.9
- BBHDeepSeek-V3 by 5.187.582.4+5.1
- MBPPQwen2 Instruct (72B) by 4.875.480.2+4.8
- AA IntelligenceDeepSeek-V3 by 3.49.76.3+3.4
- simple_safety_testsQwen2 Instruct (72B) by 3.295.398.5+3.2
- GSM8KDeepSeek-V3 by 2.99491.1+2.9
- C-EvalDeepSeek-V3 by 2.786.583.8+2.7
- anthropic_red_teamQwen2 Instruct (72B) by 297.199.1+2
- bbqDeepSeek-V3 by 1.696.795.1+1.6
- CMMLUQwen2 Instruct (72B) by 1.388.890.1+1.3
- HellaSwagDeepSeek-V3 by 1.388.987.6+1.3
- WinoGrandetie84.985.1tie
- XSTesttie97.196.9tie
- OpenBookQAtie95.495.4tie
Questions people ask
Which is better, DeepSeek-V3 or Qwen2 Instruct (72B)?
DeepSeek-V3 wins all two areas where both have results: coding and reasoning. Qwen2 Instruct (72B) wins none.
Which is better for coding?
DeepSeek-V3. It wins the one coding test both models report.
How do you compare the two?
We use the 26 benchmark tests both models have published scores on. The verdict counts the 3 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 23 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.