Phi 4 Multimodal Instruct vs Qwen3 VL 4B (Reasoning)
Wins 0 of 3 areas
—
Wins 3 of 3 areas
Coding · Reasoning · Images and charts
Qwen3 VL 4B (Reasoning) is the stronger all-rounder.
Scores updated · 33 tests both models report · How we compare
Where each one wins
Tests won in each of the three 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.
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
- Harder college exam questions with imagesMMMU-Pro+37.5points ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+17.9points ahead
- Math problems shown in pictures and chartsMathVista+17.1points ahead
Where Phi 4 Multimodal Instruct pulls ahead
- Reads text in imagesOCRBench+3.6points ahead
Every test, side by side
All 33 tests both models report. The winning score is in its model's colour; marks a score checked independently.
ReasoningQwen3 VL 4B (Reasoning)
- GPQA DiamondQwen3 VL 4B (Reasoning) by 17.931.549.4+17.9
- Humanity's Last Examtie54.6tie
Images and chartsQwen3 VL 4B (Reasoning)
Other results24 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.
- TextVQA-valQwen3 VL 4B (Reasoning) by 40.739.980.5+40.7
- MultiDocVQA-valQwen3 VL 4B (Reasoning) by 40.446.887.2+40.4
- VSI-BenchQwen3 VL 4B (Reasoning) by 31.124.155.2+31.1
- MME-RealWorldQwen3 VL 4B (Reasoning) by 30.732.563.2+30.7
- MME-PerceptionQwen3 VL 4B (Reasoning) by 294 rating points14101704+294 rating
- CV-BenchQwen3 VL 4B (Reasoning) by 28.657.185.7+28.6
- LVBenchQwen3 VL 4B (Reasoning) by 28.225.353.5+28.2
- Timelens-ActivityNetQwen3 VL 4B (Reasoning) by 26.4228.4+26.4
- NextQAQwen3 VL 4B (Reasoning) by 25.754.179.8+25.7
- MV-BenchQwen3 VL 4B (Reasoning) by 24.444.969.3+24.4
- MMBench-EN-devQwen3 VL 4B (Reasoning) by 17.565.883.3+17.5
- MLVU-devQwen3 VL 4B (Reasoning) by 17.344.261.5+17.3
- LongVideoBenchQwen3 VL 4B (Reasoning) by 16.641.157.7+16.6
- InfoVQA (val)Qwen3 VL 4B (Reasoning) by 7.771.879.5+7.7
- TextVQAQwen3 VL 4B (Reasoning) by 6.275.681.8+6.2
- MMBench-CN-devQwen3 VL 4B (Reasoning) by 5.475.280.6+5.4
- Ref-DAVIS17Qwen3 VL 4B (Reasoning) by 4.43.17.5+4.4
- AI2DQwen3 VL 4B (Reasoning) by 2.682.384.9+2.6
- ChartQAQwen3 VL 4B (Reasoning) by 1.981.483.3+1.9
- DocVQA-valQwen3 VL 4B (Reasoning) by 1.992.894.7+1.9
- DocVQAQwen3 VL 4B (Reasoning) by 1.793.294.9+1.7
- SATQwen3 VL 4B (Reasoning) by 1.455.356.7+1.4
- AA IntelligenceQwen3 VL 4B (Reasoning) by 1.25.87+1.2
- AI2D w/ Masktie81.881.5tie
Questions people ask
Which is better, Phi 4 Multimodal Instruct or Qwen3 VL 4B (Reasoning)?
Qwen3 VL 4B (Reasoning) wins all three areas where both have results: coding, reasoning and images and charts. Phi 4 Multimodal Instruct wins none.
Which is better for coding?
Qwen3 VL 4B (Reasoning). It wins the one coding test both models report.
How do you compare the two?
We use the 33 benchmark tests both models have published scores on. The verdict counts the 9 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 24 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.