Qwen2.5 VL 72B vs Qwen3 VL 4B (Reasoning)
Wins 0 of 2 areas
—
Wins 2 of 2 areas
Agents · Images and charts
Qwen3 VL 4B (Reasoning) is the stronger all-rounder.
Scores updated · 22 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 Qwen3 VL 4B (Reasoning) pulls ahead
- Completes tasks by operating a computer desktopOSWorld-Verified+22.6points ahead
- Math problems shown in pictures and chartsMathVista+4.7points ahead
- Questions that truly need the image to answerMMStar+2.4points ahead
Where Qwen2.5 VL 72B pulls ahead
- Reads text in imagesOCRBench+7.7points ahead
Every test, side by side
All 22 tests both models report. The winning score is in its model's colour; marks a score checked independently.
AgentsQwen3 VL 4B (Reasoning)
- OSWorld-VerifiedQwen3 VL 4B (Reasoning) by 22.68.831.4+22.6
Images and chartsQwen3 VL 4B (Reasoning)
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.
- MATH-VisionQwen3 VL 4B (Reasoning) by 21.938.160+21.9
- ChartQAProQwen2.5 VL 72B by 9.145.336.2+9.1
- HallusionBenchQwen3 VL 4B (Reasoning) by 8.955.264.1+8.9
- ChartQAQwen2.5 VL 72B by 6.289.583.3+6.2
- LVBenchQwen3 VL 4B (Reasoning) by 6.247.353.5+6.2
- CC-OCRQwen2.5 VL 72B by 679.873.8+6
- ScreenSpotQwen3 VL 4B (Reasoning) by 5.887.192.9+5.8
- ScreenSpot-Pro (No tools)Qwen3 VL 4B (Reasoning) by 5.643.649.2+5.6
- MMMU (val) (Pass@1)Qwen3 VL 4B (Reasoning) by 565.870.8+5
- AI2DQwen2.5 VL 72B by 3.588.484.9+3.5
- TextVQA-valQwen2.5 VL 72B by 2.783.380.5+2.7
- DocVQAQwen2.5 VL 72B by 1.596.494.9+1.5
- DocVQA-valQwen2.5 VL 72B by 1.195.894.7+1.1
- MV-BenchQwen2.5 VL 72B by 1.170.469.3+1.1
- RealWorldQAtie73.973.2tie
- OCRBench v2_entie61.561.8tie
Questions people ask
Which is better, Qwen2.5 VL 72B or Qwen3 VL 4B (Reasoning)?
Qwen3 VL 4B (Reasoning) wins all two areas where both have results: agents and images and charts. Qwen2.5 VL 72B wins none.
How do you compare the two?
We use the 22 benchmark tests both models have published scores on. The verdict counts the 6 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.