← All comparisons
Qwen3.6 27B (Non-reasoning) vs Qwen3 Omni 30B A3B (Reasoning)
Alibaba vs Alibaba — side-by-side benchmark comparison
| Qwen3.6 27B (Non-reasoning) | Qwen3 Omni 30B A3B (Reasoning) | |
|---|---|---|
| Intelligence Index | 37.1 | 15.6 |
| Coding Index | 26.6 | 12.7 |
| Math Index | — | 74.0 |
| Output speed (tok/s) | 60.7 | 100.1 |
| Blended price ($/1M) | $1.35 | $0.43 |
| Time to first token (s) | 1.55s | 0.97s |
| aime | — | — |
| aime 25 | — | 74.0% |
| artificial analysis coding index | 26.60 | 12.70 |
| artificial analysis intelligence index | 37.10 | 15.60 |
| artificial analysis math index | — | 74.00 |
| gpqa | 82.9% | 72.6% |
| hle | 13.6% | 7.3% |
| ifbench | 45.7% | 43.4% |
| lcr | 55.0% | 0.0% |
| livecodebench | — | 67.9% |
| math 500 | — | — |
| mmlu pro | — | 79.2% |
| scicode | 37.3% | 30.6% |
| tau2 | 93.6% | 21.3% |
| terminalbench hard | 21.2% | 3.8% |
Benchmark data from Artificial Analysis.