← All comparisons
NVIDIA Nemotron Nano 12B v2 VL (Non-reasoning) vs Qwen3 VL 8B (Reasoning)
NVIDIA vs Alibaba — side-by-side benchmark comparison
| NVIDIA Nemotron Nano 12B v2 VL (Non-reasoning) | Qwen3 VL 8B (Reasoning) | |
|---|---|---|
| Intelligence Index | 10.1 | 16.7 |
| Coding Index | 5.9 | 9.8 |
| Math Index | 26.7 | 30.7 |
| Output speed (tok/s) | 227.4 | 133.6 |
| Blended price ($/1M) | $0.30 | $0.66 |
| Time to first token (s) | 0.63s | 1.13s |
| aime | — | — |
| aime 25 | 26.7% | 30.7% |
| artificial analysis coding index | 5.90 | 9.80 |
| artificial analysis intelligence index | 10.10 | 16.70 |
| artificial analysis math index | 26.70 | 30.70 |
| gpqa | 43.9% | 57.9% |
| hle | 4.5% | 3.3% |
| ifbench | 25.9% | 39.9% |
| lcr | 17.0% | 31.0% |
| livecodebench | 34.5% | 35.3% |
| math 500 | — | — |
| mmlu pro | 64.9% | 74.9% |
| scicode | 17.6% | 21.9% |
| tau2 | 19.3% | 22.5% |
| terminalbench hard | 0.0% | 3.8% |
Benchmark data from Artificial Analysis.