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 InternVL3_5-4B 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.
- Math & logical reasoningLFM3-VL-2B by 22.446.468.8+22.4
- MM-IFEvalLFM3-VL-2B by 18.354.572.8+18.3
- OCRBench v2_enLFM3-VL-2B by 14.449.163.5+14.4
- HallusionBenchLFM3-VL-2B by 13.552.165.6+13.5
- ScreenSpot-v2 DesktopLFM3-VL-2B by 12.68294.6+12.6
- PerceptionLFM3-VL-2B by 12.463.275.6+12.4
- Document, OCR & ChartLFM3-VL-2B by 11.872.183.9+11.8
- RealWorldQALFM3-VL-2B by 11.667.779.3+11.6
- TextVQA-valLFM3-VL-2B by 10.477.587.9+10.4
- InfographicVQA (val)LFM3-VL-2B by 9.376.986.2+9.3
- ScreenSpot-v2 WebLFM3-VL-2B by 7.382.689.9+7.3
- Multilingual MMBenchLFM3-VL-2B by 776.683.6+7
- ChartQA TestLFM3-VL-2B by 6.586.292.7+6.5
- MMELFM3-VL-2B by 5.98186.9+5.9
- MMBenchLFM3-VL-2B by 581.186.1+5
- MMMU (val) (Pass@1)InternVL3_5-4B by 560.755.7+5
- DocVQA-valLFM3-VL-2B by 4.991.896.7+4.9
- GeneralLFM3-VL-2B by 4.861.266+4.8
- ScreenSpot-v2 MobileLFM3-VL-2B by 4.487.892.2+4.4
- OCRBench v1LFM3-VL-2B by 4.38286.3+4.3
- MMMBLFM3-VL-2B by 3.781.585.2+3.7
- SimpleVQALFM3-VL-2B by 2.633.736.3+2.6
- SEED-Bench (image)LFM3-VL-2B by 2.276.478.6+2.2
- POPEtie88.988.7tie
Questions people ask
Which is better, InternVL3_5-4B or LFM3-VL-2B?
LFM3-VL-2B wins the one area where both have results: images and charts. InternVL3_5-4B 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.