Qwen2.5 VL 32B Instruct vs Qwen2.5 VL 72B
Wins 0 of 2 areas
—
Wins 2 of 2 areas
Agents · Images and charts
Qwen2.5 VL 72B is the stronger all-rounder.
Scores updated · 16 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.
Agents 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 Qwen2.5 VL 72B pulls ahead
- Completes tasks by operating a computer desktopOSWorld-Verified+2.9points ahead
- Harder college exam questions with imagesMMMU-Pro+1.6points ahead
- Questions that truly need the image to answerMMStar+1.3points ahead
Where Qwen2.5 VL 32B Instruct pulls ahead
No clear win on a test scored out of 100.
Every test, side by side
All 16 tests both models report. The winning score is in its model's colour; marks a score checked independently.
Images and chartsQwen2.5 VL 72B
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.
- OCRBench v2_enQwen2.5 VL 72B by 4.357.261.5+4.3
- ScreenSpot-Pro (Acc)Qwen2.5 VL 72B by 4.239.443.6+4.2
- ScreenSpot-Pro (No tools)Qwen2.5 VL 72B by 4.239.443.6+4.2
- CC-OCRQwen2.5 VL 72B by 2.777.179.8+2.7
- LVBenchQwen2.5 VL 32B Instruct by 1.74947.3+1.7
- DocVQAQwen2.5 VL 72B by 1.694.896.4+1.6
- ScreenSpotQwen2.5 VL 32B Instruct by 1.488.587.1+1.4
- MMStar (Acc)Qwen2.5 VL 72B by 1.369.570.8+1.3
- VideoMME (w sub.)Qwen2.5 VL 72B by 1.277.979.1+1.2
- MATH-Visiontie38.438.1tie
- MMBench-Videotie1.92tie
Questions people ask
Which is better, Qwen2.5 VL 32B Instruct or Qwen2.5 VL 72B?
Qwen2.5 VL 72B wins all two areas where both have results: agents and images and charts. Qwen2.5 VL 32B Instruct wins none.
How do you compare the two?
We use the 16 benchmark tests both models have published scores on. The verdict counts the 5 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.