| instruction-following-on-ifbench | LFM2.5-1.2B-Instruct | – | Accuracy: 47.33 |
| instruction-following-on-multi-if | LFM2.5-1.2B-Instruct | – | Accuracy: 60.98 |
| mathematical-reasoning-on-aime-2025 | LFM2.5-1.2B-Instruct | – | Accuracy: 14.00 |
| multi-task-language-understanding-on-mmlu-pro | LFM2.5-1.2B-Instruct | – | Accuracy: 44.35 |
| multimodal-reasoning-on-mmmu | LFM2.5-VL-1.6B | – | Accuracy: 40.56 |
| speech-recognition-on-librispeech-test-clean | LFM2.5-Audio-1.5B | – | Word Error Rate (WER): 1.95 |
| speech-recognition-on-librispeech-test-other | LFM2.5-Audio-1.5B | – | Word Error Rate (WER): 4.30 |
| speech-recognition-on-spgispeech | LFM2.5-Audio-1.5B | – | Word Error Rate (WER): 2.76 |
| vision-language-understanding-on-mmstar | LFM2.5-VL-1.6B | – | Accuracy: 50.67 |
| visual-question-answering-vqa-on-blink | LFM2.5-VL-1.6B | – | Accuracy: 48.82 |
| visual-question-answering-vqa-on-realworldqa | LFM2.5-VL-1.6B | – | Accuracy: 64.84 |