Qwen2 Instruct (72B) is the stronger all-rounder.
Scores updated · 17 tests both models report · How we compare
Where each one wins
Tests won in each of the one area 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.
Reasoning rests on a single test.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where Qwen2 Instruct (72B) pulls ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+12.4points ahead
Where Mistral 7B pulls ahead
No clear win on a test scored out of 100.
Every test, side by side
All 17 tests both models report. The winning score is in its model's colour; marks a score checked independently.
ReasoningQwen2 Instruct (72B)
- GPQA DiamondQwen2 Instruct (72B) by 12.424.737.1+12.4
Other results16 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.
- MATHQwen2 Instruct (72B) by 65.913.179+65.9
- HumanEvalQwen2 Instruct (72B) by 56.729.386+56.7
- EvalPlusQwen2 Instruct (72B) by 42.636.479+42.6
- MultiPL-EQwen2 Instruct (72B) by 39.829.469.2+39.8
- GSM8KQwen2 Instruct (72B) by 38.952.291.1+38.9
- C-EvalQwen2 Instruct (72B) by 36.447.483.8+36.4
- MMLU-ProQwen2 Instruct (72B) by 33.530.964.4+33.5
- MBPPQwen2 Instruct (72B) by 29.151.180.2+29.1
- BBHQwen2 Instruct (72B) by 26.356.182.4+26.3
- Theorem QAQwen2 Instruct (72B) by 25.219.244.4+25.2
- OpenBookQAQwen2 Instruct (72B) by 15.679.895.4+15.6
- MMLUQwen2 Instruct (72B) by 12.764.276.9+12.7
- TruthfulQAQwen2 Instruct (72B) by 12.642.254.8+12.6
- ARC-ChallengeQwen2 Instruct (72B) by 8.96068.9+8.9
- WinoGrandeQwen2 Instruct (72B) by 6.778.485.1+6.7
- HellaSwagQwen2 Instruct (72B) by 4.483.287.6+4.4
Questions people ask
Which is better, Mistral 7B or Qwen2 Instruct (72B)?
Qwen2 Instruct (72B) wins the one area where both have results: reasoning. Mistral 7B wins none.
How do you compare the two?
We use the 17 benchmark tests both models have published scores on. The verdict counts the 1 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 16 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.