Other · model
LFM2.5-VL-3B benchmark scores
As of 2026-10-07, LFM2.5-VL-3B (Other) has measured scores on 8 benchmarks and estimated scores on 17 more.
| Benchmark | Score | Source |
|---|---|---|
| GDPval-AA | 0.0% | estimated ± 5.8 pp, low confidence |
| IFBench | 25.8% | measured |
| IFEval | 82.3% | measured |
| τ²-bench results | 55.1% | estimated ± 14.1 pp, low confidence |
| AA-MMMU-Pro | 16.2% | estimated ± 1.2 pp, medium confidence |
| BFCL v4 | 32.5% | measured |
| AA-IFBench | 26.5% | estimated ± 8.2 pp, low confidence |
| CharXiv | 41.4% | estimated ± 4.1 pp, low confidence |
| CharXiv w/o tools | 30.8% | estimated ± 1.8 pp, low confidence |
| Claw-Eval | 28.1% | estimated ± 2.3 pp, low confidence |
| MathVision | 24.3% | estimated ± 3.6 pp, medium confidence |
| MMMU-Pro | 30.5% | measured |
| ScreenSpot Pro | 85.1% | estimated ± 2.5 pp, low confidence |
| V* | 68.2% | estimated ± 1.4 pp, low confidence |
| VideoMMMU | 84.1% | estimated ± 1.0 pp, medium confidence |
| ERQA | 39.7% | estimated ± 3.2 pp, low confidence |
| MedXpertQA (MM) | 0.7% | estimated ± 2.4 pp, medium confidence |
| MMMU | 48.4% | measured |
| SimpleVQA | 35.4% | measured |
| ZeroBench | 74.8% | estimated ± 6.4 pp, low confidence |
| MMMU-Pro w/ Python | 60.9% | estimated ± 0.6 pp, medium confidence |
| RefCOCO (avg) | 87.9% | measured |
| RealWorldQA | 73.1% | measured |
| Video-MME (with subtitle) | 80.4% | estimated ± 1.8 pp, low confidence |
| OmniDocBench 1.5 | 0.0% | estimated ± 5.9 pp, low confidence |