← All comparisons
Nemotron 3 Nano Omni 30B A3B Reasoning vs Qwen3 235B A22B (Reasoning)
NVIDIA vs Alibaba — side-by-side benchmark comparison
| Nemotron 3 Nano Omni 30B A3B Reasoning | Qwen3 235B A22B (Reasoning) | |
|---|---|---|
| Intelligence Index | 21.4 | 19.8 |
| Coding Index | 14.8 | 17.4 |
| Math Index | — | 82.0 |
| Output speed (tok/s) | 285.9 | 58.3 |
| Blended price ($/1M) | $0.13 | $2.63 |
| Time to first token (s) | 0.58s | 1.37s |
| aime | — | 84.0% |
| aime 25 | — | 82.0% |
| artificial analysis coding index | 14.80 | 17.40 |
| artificial analysis intelligence index | 21.40 | 19.80 |
| artificial analysis math index | — | 82.00 |
| gpqa | 46.9% | 70.0% |
| hle | 5.3% | 11.7% |
| ifbench | 63.2% | 38.7% |
| lcr | 35.7% | 0.0% |
| livecodebench | — | 62.2% |
| math 500 | — | 93.0% |
| mmlu pro | — | 82.8% |
| scicode | 27.8% | 39.9% |
| tau2 | 45.3% | 24.0% |
| terminalbench hard | 8.3% | 6.1% |
Benchmark data from Artificial Analysis.