← All comparisons
Qwen3 Next 80B A3B (Reasoning) vs Llama 3.3 Nemotron Super 49B v1 (Reasoning)
Alibaba vs NVIDIA — side-by-side benchmark comparison
| Qwen3 Next 80B A3B (Reasoning) | Llama 3.3 Nemotron Super 49B v1 (Reasoning) | |
|---|---|---|
| Intelligence Index | 26.7 | 18.5 |
| Coding Index | 19.5 | 9.4 |
| Math Index | 84.3 | 54.7 |
| Output speed (tok/s) | 149.4 | 0.0 |
| Blended price ($/1M) | $1.88 | $0.00 |
| Time to first token (s) | 1.23s | 0.00s |
| aime | — | 58.3% |
| aime 25 | 84.3% | 54.7% |
| artificial analysis coding index | 19.50 | 9.40 |
| artificial analysis intelligence index | 26.70 | 18.50 |
| artificial analysis math index | 84.30 | 54.70 |
| gpqa | 75.9% | 64.3% |
| hle | 11.7% | 6.5% |
| ifbench | 60.7% | 38.1% |
| lcr | 60.3% | 17.0% |
| livecodebench | 78.4% | 27.7% |
| math 500 | — | 95.9% |
| mmlu pro | 82.4% | 78.5% |
| scicode | 38.8% | 28.2% |
| tau2 | 41.5% | 26.9% |
| terminalbench hard | 9.8% | 0.0% |
Benchmark data from Artificial Analysis.