Qwen2 Instruct (72B) is the stronger all-rounder.
Scores updated · 19 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+11.4points ahead
Where Gemma (7B) pulls ahead
No clear win on a test scored out of 100.
Every test, side by side
All 19 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 11.425.737.1+11.4
Other results18 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.
- HumanEvalQwen2 Instruct (72B) by 48.837.286+48.8
- C-EvalQwen2 Instruct (72B) by 40.243.683.8+40.2
- MultiPL-EQwen2 Instruct (72B) by 39.529.769.2+39.5
- EvalPlusQwen2 Instruct (72B) by 39.439.679+39.4
- GSM8KQwen2 Instruct (72B) by 35.255.991.1+35.2
- MMLU-ProQwen2 Instruct (72B) by 30.733.764.4+30.7
- MBPPQwen2 Instruct (72B) by 29.650.680.2+29.6
- MATHQwen2 Instruct (72B) by 295079+29
- BBHQwen2 Instruct (72B) by 27.355.182.4+27.3
- Theorem QAQwen2 Instruct (72B) by 22.921.544.4+22.9
- MMLUQwen2 Instruct (72B) by 19.857.176.9+19.8
- OpenBookQAQwen2 Instruct (72B) by 14.680.895.4+14.6
- TruthfulQAQwen2 Instruct (72B) by 1044.854.8+10
- ARC-ChallengeQwen2 Instruct (72B) by 7.861.168.9+7.8
- WinoGrandeQwen2 Instruct (72B) by 6.17985.1+6.1
- HellaSwagQwen2 Instruct (72B) by 5.482.287.6+5.4
- naturalquestions_closedbookQwen2 Instruct (72B) by 5.433.639+5.4
- NarrativeQAGemma (7B) by 2.575.272.7+2.5
Questions people ask
Which is better, Gemma (7B) or Qwen2 Instruct (72B)?
Qwen2 Instruct (72B) wins the one area where both have results: reasoning. Gemma (7B) wins none.
How do you compare the two?
We use the 19 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 18 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.