| Benchmark | Model | Rank | Results |
|---|---|---|---|
| code-generation-on-humaneval | MELT-1.6B | – | Pass@1: 81.7 |
| mathematical-reasoning-on-aime-2025 | MELT-1.6B | – | Accuracy: 33.3 |
| mathematical-reasoning-on-aime-2026 | MELT-1.6B | – | Accuracy: 41.0 |
| mathematical-reasoning-on-aime24 | MELT-1.6B | – | Acc: 46.7 |
| mathematical-reasoning-on-math-500 | MELT-1.6B | – | Accuracy: 93.4 |
| mathematical-reasoning-on-olympiadbench | MELT-1.6B | – | Accuracy: 64.7 |
| question-answering-on-humanity-s-last-exam-hle | MELT-1.6B | – | Accuracy: 2.0 |