LFM3-VL-2B is the stronger all-rounder.
Scores updated · 28 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.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where LFM3-VL-2B pulls ahead
Where Qwen3.5 2B pulls ahead
No clear win on a test scored out of 100.
Every test, side by side
All 28 tests both models report. The winning score is in its model's colour; marks a score checked independently.
Images and chartsLFM3-VL-2B
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.
- ScreenSpot-v2 DesktopLFM3-VL-2B by 30.894.663.8+30.8
- ScreenSpot-v2 WebLFM3-VL-2B by 2489.965.9+24
- ScreenSpot-v2 MobileLFM3-VL-2B by 22.592.269.7+22.5
- MM-IFEvalLFM3-VL-2B by 17.472.855.4+17.4
- OCRBench v2_enLFM3-VL-2B by 15.463.548.1+15.4
- ChartQA TestLFM3-VL-2B by 14.392.778.4+14.3
- RealWorldQALFM3-VL-2B by 14.279.365.1+14.2
- Multilingual MMBenchLFM3-VL-2B by 13.783.669.9+13.7
- MMBenchLFM3-VL-2B by 1386.173.1+13
- InfographicVQA (val)LFM3-VL-2B by 12.786.273.5+12.7
- MMMU (val) (Pass@1)LFM3-VL-2B by 11.655.744.1+11.6
- MMELFM3-VL-2B by 10.786.976.2+10.7
- MMMBLFM3-VL-2B by 10.785.274.5+10.7
- TextVQA-valLFM3-VL-2B by 10.687.977.3+10.6
- Math & logical reasoningLFM3-VL-2B by 7.868.861+7.8
- Document, OCR & ChartLFM3-VL-2B by 4.883.979.1+4.8
- PerceptionLFM3-VL-2B by 4.475.671.2+4.4
- DocVQA-valLFM3-VL-2B by 4.196.792.6+4.1
- SEED-Bench (image)LFM3-VL-2B by 2.878.675.8+2.8
- GeneralLFM3-VL-2B by 26664+2
- OCRBench v1LFM3-VL-2B by 1.986.384.4+1.9
- SimpleVQALFM3-VL-2B by 1.136.335.2+1.1
- HallusionBenchtie65.665.5tie
- POPEtie88.788.6tie
Questions people ask
Which is better, LFM3-VL-2B or Qwen3.5 2B?
LFM3-VL-2B wins the one area where both have results: images and charts. Qwen3.5 2B wins none.
How do you compare the two?
We use the 28 benchmark tests both models have published scores on. The verdict counts the 4 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.