# LFM3-VL-2B benchmark scores

Other · model · https://benchgap.net/model/lfm3-vl-2b · data retrieved 2026-10-29

As of 2026-10-29, LFM3-VL-2B (Other) has measured scores on 20 benchmarks and estimated scores on 7 more.

| Benchmark | Score | Source |
|---|---|---|
| [AA-MMMU-Pro](https://benchgap.net/b/aa-mmmu-pro/current.md) | 45.4% | estimated ± 1.6 pp, high confidence |
| [MMMU](https://benchgap.net/b/mmmu/current.md) | 55.7% | measured |
| [ParseBench](https://benchgap.net/b/parsebench/current.md) | 61.1% | estimated ± 0.8 pp, low confidence |
| [CharXiv](https://benchgap.net/b/charxiv/current.md) | 67.6% | estimated ± 1.6 pp, low confidence |
| [CharXiv w/o tools](https://benchgap.net/b/charxivnotools/current.md) | 70.2% | measured |
| [AI2D_TEST](https://benchgap.net/b/ai2dtest/current.md) | 82.7% | estimated ± 1.3 pp, low confidence |
| [CountBench](https://benchgap.net/b/countbench/current.md) | 93.0% | measured |
| [DynaMath](https://benchgap.net/b/dynamath/current.md) | 79.2% | measured |
| [ERQA](https://benchgap.net/b/erqa/current.md) | 48.2% | measured |
| [MathVision](https://benchgap.net/b/mathvision/current.md) | 51.6% | measured |
| [MLVU (M-Avg)](https://benchgap.net/b/mlvuavg/current.md) | 71.3% | measured |
| [MMMU-Pro](https://benchgap.net/b/mmmu-pro/current.md) | 45.7% | measured |
| [OmniDocBench 1.5](https://benchgap.net/b/omnidocbench15/current.md) | 87.6% | estimated ± 1.8 pp, medium confidence |
| [RealWorldQA](https://benchgap.net/b/realworldqa/current.md) | 79.3% | measured |
| [ScreenSpot Pro](https://benchgap.net/b/screenspot-pro/current.md) | 55.5% | measured |
| [SimpleVQA](https://benchgap.net/b/simplevqa/current.md) | 36.3% | measured |
| [V*](https://benchgap.net/b/vstar/current.md) | 86.4% | measured |
| [Video-MME (w/o subtitle)](https://benchgap.net/b/videommenosub/current.md) | 62.8% | measured |
| [Video-MME (with subtitle)](https://benchgap.net/b/videommewithsub/current.md) | 69.0% | measured |
| [VideoMMMU](https://benchgap.net/b/videommmu/current.md) | 49.6% | measured |
| [We-Math](https://benchgap.net/b/wemath/current.md) | 68.6% | measured |
| [MedXpertQA (MM)](https://benchgap.net/b/medxpertqamm/current.md) | 58.1% | estimated ± 4.2 pp, low confidence |
| [MMVU](https://benchgap.net/b/mmvu/current.md) | 52.9% | measured |
| [LVBench](https://benchgap.net/b/lvbench/current.md) | 41.2% | measured |
| [CC-OCR](https://benchgap.net/b/ccocr/current.md) | 81.0% | estimated ± 0.7 pp, medium confidence |
| [RefCOCO (avg)](https://benchgap.net/b/refcocoavg/current.md) | 90.5% | measured |
| [OCRBench V2](https://benchgap.net/b/ocrbenchv2/current.md) | 63.5% | measured |