Other · model
Mellum2.1-12B-A2.5B-Thinking benchmark scores
As of 2026-10-09, Mellum2.1-12B-A2.5B-Thinking (Other) has measured scores on 7 benchmarks and estimated scores on 43 more.
| Benchmark | Score | Source |
|---|---|---|
| AA-GPQA Diamond | 62.9% | estimated ± 5.6 pp, medium confidence |
| AA-HLE | 6.6% | estimated ± 4.0 pp, high confidence |
| AA-SciCode | 40.9% | estimated ± 3.7 pp, high confidence |
| Artificial Analysis Intelligence Index | 16.7% | estimated ± 4.2 pp, medium confidence |
| CritPt | 0.0% | estimated ± 4.9 pp, high confidence |
| GPQA Diamond | 64.6% | measured |
| HLE | 15.9% | estimated ± 5.9 pp, low confidence |
| IFBench | 64.8% | estimated ± 9.7 pp, medium confidence |
| LiveCodeBench v6 | 82.0% | measured |
| AA-Omniscience Accuracy | 10.3% | estimated ± 4.5 pp, medium confidence |
| SciCode | 31.7% | estimated ± 3.8 pp, high confidence |
| Terminal-Bench 2.1 | 17.4% | measured |
| IFEval | 90.6% | measured |
| AA Coding Index | 23.3% | estimated ± 5.1 pp, medium confidence |
| SWE-bench Verified | 47.0% | measured |
| SWE-bench Pro | 28.0% | measured |
| MMLU-Pro | 75.7% | estimated ± 1.4 pp, high confidence |
| GPQA | 64.9% | estimated ± 0.1 pp, high confidence |
| AA-IFBench | 53.5% | estimated ± 8.2 pp, medium confidence |
| ARC-AGI-1 | 33.9% | estimated ± 2.7 pp, medium confidence |
| ARC-AGI-2 | 15.2% | estimated ± 10.1 pp, low confidence |
| CursorBench 3.1 | 29.8% | estimated ± 5.6 pp, low confidence |
| CursorBench 3.2 | 2.6% | estimated ± 3.6 pp, medium confidence |
| Terminal-Bench 3.0 | 0.8% | estimated ± 9.1 pp, low confidence |
| FrontierCode 1.1 Main | 0.0% | estimated ± 2.7 pp, low confidence |
| FrontierSWE v2 | 0.0% | estimated ± 9.0 pp, low confidence |
| Vals GPQA Diamond | 63.2% | estimated ± 2.0 pp, medium confidence |
| Vals LiveCodeBench | 14.6% | estimated ± 4.7 pp, low confidence |
| Vals MMLU-Pro | 70.8% | estimated ± 1.6 pp, medium confidence |
| Vals SWE-bench | 31.0% | estimated ± 6.8 pp, low confidence |
| Terminal-Bench 2.1 (Vals AI) | 19.4% | estimated ± 7.4 pp, low confidence |
| AA Terminal-Bench 2.1 | 52.1% | estimated ± 3.5 pp, medium confidence |
| AA Terminal-Bench 4.0 | 0.0% | estimated ± 8.9 pp, low confidence |
| CursorBench 4.0 | 11.0% | estimated ± 3.6 pp, medium confidence |
| HLE w/o tools | 5.8% | estimated ± 4.4 pp, medium confidence |
| PostTrainBench v1.1 | 9.7% | estimated ± 4.7 pp, medium confidence |
| Vibe Code Bench | 0.6% | estimated ± 8.1 pp, low confidence |
| MMLU-Redux | 87.8% | measured |
| NL2Repo | 43.2% | estimated ± 3.1 pp, low confidence |
| SuperGPQA | 55.6% | estimated ± 4.5 pp, high confidence |
| LiveCodeBench Pro | 51.2% | estimated ± 6.0 pp, low confidence |
| MedXpertQA (Text) | 18.3% | estimated ± 4.5 pp, low confidence |
| MMLU-Pro (Arcee) | 76.4% | estimated ± 3.4 pp, medium confidence |
| SWE-Rebench | 57.0% | estimated ± 3.6 pp, low confidence |
| ARC-AGI-3 | 0.0% | estimated ± 3.8 pp, medium confidence |
| FrontierCode 1.1 Extended | 39.9% | estimated ± 3.0 pp, low confidence |
| HealthBench Professional | 10.7% | estimated ± 4.8 pp, low confidence |
| LiveCodeBench | 61.8% | estimated ± 10.2 pp, low confidence |
| OpenHarmony Bench | 53.3% | estimated ± 4.9 pp, low confidence |
| MMMLU | 65.4% | estimated ± 1.1 pp, low confidence |