LFM3-VL-2B is the stronger all-rounder.
Scores updated · 24 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_5-2B pulls ahead
No clear win on a test scored out of 100.
Every test, side by side
All 24 tests both models report. The winning score is in its model's colour; marks a score checked independently.
Images and chartsLFM3-VL-2B
Other results20 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.
- MM-IFEvalLFM3-VL-2B by 25.747.172.8+25.7
- HallusionBenchLFM3-VL-2B by 1847.665.6+18
- OCRBench v2_enLFM3-VL-2B by 1845.563.5+18
- RealWorldQALFM3-VL-2B by 17.761.679.3+17.7
- InfographicVQA (val)LFM3-VL-2B by 16.969.386.2+16.9
- ScreenSpot-v2 DesktopLFM3-VL-2B by 14.779.994.6+14.7
- MMELFM3-VL-2B by 13.373.686.9+13.3
- Multilingual MMBenchLFM3-VL-2B by 12.770.983.6+12.7
- TextVQA-valLFM3-VL-2B by 11.376.687.9+11.3
- ChartQA TestLFM3-VL-2B by 1181.792.7+11
- ScreenSpot-v2 WebLFM3-VL-2B by 1079.989.9+10
- MMBenchLFM3-VL-2B by 9.976.286.1+9.9
- MMMBLFM3-VL-2B by 8.976.385.2+8.9
- DocVQA-valLFM3-VL-2B by 8.388.496.7+8.3
- ScreenSpot-v2 MobileLFM3-VL-2B by 686.292.2+6
- SimpleVQALFM3-VL-2B by 5.830.536.3+5.8
- MMMU (val) (Pass@1)LFM3-VL-2B by 3.75255.7+3.7
- SEED-Bench (image)LFM3-VL-2B by 3.275.478.6+3.2
- OCRBench v1LFM3-VL-2B by 2.483.986.3+2.4
- POPEtie8888.7tie
Questions people ask
Which is better, InternVL3_5-2B or LFM3-VL-2B?
LFM3-VL-2B wins the one area where both have results: images and charts. InternVL3_5-2B wins none.
How do you compare the two?
We use the 24 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 20 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.