Other · model
Ternary Bonsai 2 27B benchmark scores
As of 2026-10-07, Ternary Bonsai 2 27B (Other) has measured scores on 14 benchmarks and estimated scores on 54 more.
| Benchmark | Score | Source |
|---|---|---|
| AA-GPQA Diamond | 84.3% | estimated ± 5.4 pp, medium confidence |
| AA-HLE | 26.8% | estimated ± 4.0 pp, high confidence |
| AA-SciCode | 47.2% | estimated ± 3.7 pp, high confidence |
| AIME26 | 95.8% | measured |
| Artificial Analysis Intelligence Index | 18.0% | estimated ± 4.2 pp, medium confidence |
| CritPt | 2.1% | estimated ± 4.1 pp, high confidence |
| GDPval-AA | 13.2% | estimated ± 12.5 pp, low confidence |
| GPQA Diamond | 85.8% | measured |
| HLE | 18.7% | estimated ± 5.9 pp, low confidence |
| IFBench | 74.0% | measured |
| AA-LCR | 77.0% | measured |
| LiveCodeBench v6 | 90.1% | measured |
| AA-Omniscience Accuracy | 11.7% | estimated ± 4.5 pp, medium confidence |
| SciCode | 39.1% | estimated ± 3.8 pp, high confidence |
| Terminal-Bench 2.1 | 52.8% | measured |
| IFEval | 91.3% | measured |
| LongBench v2 | 61.5% | estimated ± 5.9 pp, medium confidence |
| τ²-bench results | 80.2% | measured |
| AA Coding Index | 29.4% | estimated ± 5.1 pp, medium confidence |
| AA-MMMU-Pro | 67.4% | estimated ± 1.9 pp, low confidence |
| APEX-Agents-AA | 1.6% | estimated ± 11.4 pp, low confidence |
| SWE-bench Verified | 60.8% | measured |
| BFCL v4 | 56.8% | estimated ± 9.8 pp, low confidence |
| SWE-bench Pro | 34.9% | estimated ± 3.2 pp, high confidence |
| AA Agentic Index | 19.1% | estimated ± 12.3 pp, low confidence |
| MMLU-Pro | 77.9% | estimated ± 1.4 pp, high confidence |
| GPQA | 85.8% | measured |
| AA-IFBench | 57.0% | estimated ± 8.2 pp, medium confidence |
| AA EnterpriseOps-Gym | 22.9% | estimated ± 3.6 pp, low confidence |
| ARC-AGI-1 | 82.0% | estimated ± 2.7 pp, medium confidence |
| ARC-AGI-2 | 41.8% | estimated ± 10.1 pp, low confidence |
| Terminal-Bench 3.0 | 4.0% | estimated ± 9.1 pp, low confidence |
| FrontierCode 1.1 Main | 0.4% | estimated ± 2.7 pp, low confidence |
| Vals GPQA Diamond | 84.9% | estimated ± 2.0 pp, high confidence |
| Vals LiveCodeBench | 47.4% | estimated ± 9.9 pp, low confidence |
| Vals MMLU-Pro | 82.9% | estimated ± 1.2 pp, high confidence |
| Vals SWE-bench | 50.0% | estimated ± 6.8 pp, low confidence |
| Terminal-Bench 2.1 (Vals AI) | 46.7% | estimated ± 7.4 pp, low confidence |
| AA ITBench | 58.1% | estimated ± 3.6 pp, low confidence |
| AA Terminal-Bench 2.1 | 52.2% | estimated ± 3.5 pp, high confidence |
| AA Terminal-Bench 4.0 | 0.0% | estimated ± 8.9 pp, medium confidence |
| HLE w/o tools | 22.5% | estimated ± 4.4 pp, high confidence |
| Vibe Code Bench | 0.6% | estimated ± 8.1 pp, low confidence |
| CharXiv | 65.0% | estimated ± 4.1 pp, low confidence |
| C-Eval | 90.8% | estimated ± 0.9 pp, low confidence |
| HMMT Feb 2025 | 96.1% | estimated ± 1.9 pp, low confidence |
| HMMT Feb 2026 | 87.2% | estimated ± 4.0 pp, high confidence |
| MCP-Tasks | 75.7% | estimated ± 8.2 pp, low confidence |
| MMAnswerBench | 84.2% | estimated ± 1.3 pp, high confidence |
| MMLU-Redux | 89.1% | measured |
| MMMU-Pro | 61.8% | estimated ± 11.4 pp, low confidence |
| NL2Repo | 46.4% | estimated ± 3.1 pp, low confidence |
| SuperGPQA | 58.2% | estimated ± 4.5 pp, high confidence |
| ERQA | 55.4% | estimated ± 3.2 pp, low confidence |
| HealthBench Hard | 0.0% | estimated ± 5.1 pp, low confidence |
| MedXpertQA (Text) | 43.2% | estimated ± 4.5 pp, low confidence |
| MMLU-Pro (Arcee) | 85.4% | estimated ± 3.4 pp, medium confidence |
| SWE-Rebench | 57.0% | estimated ± 3.6 pp, low confidence |
| ARC-AGI-3 | 0.0% | estimated ± 2.1 pp, medium confidence |
| HealthBench Professional | 42.0% | estimated ± 4.8 pp, low confidence |
| AIME 2025 | 95.0% | measured |
| MMMU | 68.8% | estimated ± 6.6 pp, low confidence |
| LiveCodeBench | 73.3% | estimated ± 10.2 pp, low confidence |
| SimpleVQA | 54.3% | estimated ± 9.5 pp, low confidence |
| MMMLU | 88.1% | estimated ± 1.1 pp, medium confidence |
| MMLU | 90.3% | estimated ± 3.6 pp, low confidence |
| MATH-500 | 98.8% | measured |
| RealWorldQA | 80.1% | measured |