← All comparisons
NVIDIA Nemotron Nano 12B v2 VL (Reasoning) vs Qwen3.5 397B A17B (Non-reasoning)
NVIDIA vs Alibaba — side-by-side benchmark comparison
| NVIDIA Nemotron Nano 12B v2 VL (Reasoning) | Qwen3.5 397B A17B (Non-reasoning) | |
|---|---|---|
| Intelligence Index | 14.9 | 40.1 |
| Coding Index | 11.7 | 37.4 |
| Math Index | 75.0 | — |
| Output speed (tok/s) | 0.0 | 53.5 |
| Blended price ($/1M) | $0.30 | $1.35 |
| Time to first token (s) | 0.00s | 1.85s |
| aime | — | — |
| aime 25 | 75.0% | — |
| artificial analysis coding index | 11.70 | 37.40 |
| artificial analysis intelligence index | 14.90 | 40.10 |
| artificial analysis math index | 75.00 | — |
| gpqa | 57.2% | 86.1% |
| hle | 5.3% | 18.8% |
| ifbench | 31.9% | 51.6% |
| lcr | 40.0% | 58.0% |
| livecodebench | 69.4% | — |
| math 500 | — | — |
| mmlu pro | 75.9% | — |
| scicode | 26.2% | 41.1% |
| tau2 | 21.3% | 83.9% |
| terminalbench hard | 4.5% | 35.6% |
Benchmark data from Artificial Analysis.