← All comparisons
Qwen3 Next 80B A3B Instruct vs Llama 3.1 Nemotron Nano 4B v1.1 (Reasoning)
Alibaba vs NVIDIA — side-by-side benchmark comparison
| Qwen3 Next 80B A3B Instruct | Llama 3.1 Nemotron Nano 4B v1.1 (Reasoning) | |
|---|---|---|
| Intelligence Index | 20.1 | 14.4 |
| Coding Index | 15.3 | — |
| Math Index | 66.3 | 50.0 |
| Output speed (tok/s) | 141.2 | 0.0 |
| Blended price ($/1M) | $0.88 | $0.00 |
| Time to first token (s) | 1.19s | 0.00s |
| aime | — | 70.7% |
| aime 25 | 66.3% | 50.0% |
| artificial analysis coding index | 15.30 | — |
| artificial analysis intelligence index | 20.10 | 14.40 |
| artificial analysis math index | 66.30 | 50.00 |
| gpqa | 73.8% | 40.8% |
| hle | 7.3% | 5.1% |
| ifbench | 39.7% | 25.5% |
| lcr | 51.3% | 0.0% |
| livecodebench | 68.4% | 49.3% |
| math 500 | — | 94.7% |
| mmlu pro | 81.9% | 55.6% |
| scicode | 30.7% | 10.1% |
| tau2 | 21.6% | 11.7% |
| terminalbench hard | 7.6% | — |
Benchmark data from Artificial Analysis.