| Benchmark | Kimi K2.6 | Nemotron 3 Super 120B A12B |
|---|---|---|
| AA Agentic Index | 58.7 | 8.8 |
| AA Intelligence | 45.1 | 25.7 |
| AA-LCR | 76.7 | 60.3 |
| AA-Omniscience | 6.4 | -41.5 |
| aime_2026 | 96.4 | 91.7 |
| apexAgents | 28.5 | 1.8 |
| Artificial Analysis Coding Index | 61.8 | 37.7 |
| browsecomp | 86.3 | 31.3 |
| critpt | 8 | 3.1 |
| gdpval | 41.4 | 9.9 |
| GPQA Diamond | 91.1 | 82.7 |
| HLE | 37.5 | 22.8 |
| hmmt_feb_2026 | 94.7 | 85.2 |
| IFBench | 76 | 72.6 |
| itbenchSre | 31.2 | 1.1 |
| LiveCodeBench | 89.6 | 82.1 |
| MMLU-Pro | 87.1 | 83.8 |
| OmniScience Accuracy | 32.8 | 24.3 |
| OmniScience Non-Hallucination | 60.7 | 13 |
| scicode | 53.5 | 42 |
| SWE-bench Multilingual | 76.7 | 45.8 |
| SWE-bench Verified | 80.2 | 53.7 |
| TauBench V3 - Banking | 23.3 | 10.3 |
| Terminal-Bench 2.0 | 66.7 | 31 |
| Terminal-Bench 2.1 | 65.9 | 38.6 |
| Terminal-Bench Hard | 43.9 | 28.8 |
| τ²-Bench Telecom (AA run) | 95.9 | 67.8 |
| τ³-Bench | 20.6 | 10.1 |
Best tracked score per model per benchmark (default configuration; source-attributed). ↓ marks lower-is-better metrics. Open either model for its full surface, provenance and pricing.