← All comparisons
GPT-5.4 nano (xhigh) vs Qwen3 30B A3B (Non-reasoning)
OpenAI vs Alibaba — side-by-side benchmark comparison
| GPT-5.4 nano (xhigh) | Qwen3 30B A3B (Non-reasoning) | |
|---|---|---|
| Intelligence Index | 44.0 | 12.5 |
| Coding Index | 43.9 | 13.3 |
| Math Index | — | 21.7 |
| Output speed (tok/s) | 150.8 | 64.7 |
| Blended price ($/1M) | $0.46 | $0.13 |
| Time to first token (s) | 2.87s | 1.38s |
| aime | — | 26.0% |
| aime 25 | — | 21.7% |
| artificial analysis coding index | 43.90 | 13.30 |
| artificial analysis intelligence index | 44.00 | 12.50 |
| artificial analysis math index | — | 21.70 |
| gpqa | 81.7% | 51.5% |
| hle | 26.5% | 4.6% |
| ifbench | 75.9% | 31.9% |
| lcr | 66.0% | 0.0% |
| livecodebench | — | 32.2% |
| math 500 | — | 86.3% |
| mmlu pro | — | 71.0% |
| scicode | 46.9% | 26.4% |
| tau2 | 76.0% | 22.2% |
| terminalbench hard | 42.4% | 6.8% |
Benchmark data from Artificial Analysis.