← All comparisons
Qwen3 Omni 30B A3B Instruct vs GPT-5.2 (Non-reasoning)
Alibaba vs OpenAI — side-by-side benchmark comparison
| Qwen3 Omni 30B A3B Instruct | GPT-5.2 (Non-reasoning) | |
|---|---|---|
| Intelligence Index | 10.7 | 33.6 |
| Coding Index | 7.2 | 34.7 |
| Math Index | 52.3 | 51.0 |
| Output speed (tok/s) | 108.9 | 72.8 |
| Blended price ($/1M) | $0.43 | $4.81 |
| Time to first token (s) | 1.04s | 0.61s |
| aime | — | — |
| aime 25 | 52.3% | 51.0% |
| artificial analysis coding index | 7.20 | 34.70 |
| artificial analysis intelligence index | 10.70 | 33.60 |
| artificial analysis math index | 52.30 | 51.00 |
| gpqa | 62.0% | 71.2% |
| hle | 5.1% | 7.3% |
| ifbench | 31.2% | 47.4% |
| lcr | 0.0% | 38.0% |
| livecodebench | 42.2% | 66.9% |
| math 500 | — | — |
| mmlu pro | 72.5% | 81.4% |
| scicode | 18.6% | 40.4% |
| tau2 | 16.4% | 46.5% |
| terminalbench hard | 1.5% | 31.8% |
Benchmark data from Artificial Analysis.