← All comparisons
Grok 4.20 0309 v2 (Non-reasoning) vs Qwen3 30B A3B 2507 Instruct
xAI vs Alibaba — side-by-side benchmark comparison
| Grok 4.20 0309 v2 (Non-reasoning) | Qwen3 30B A3B 2507 Instruct | |
|---|---|---|
| Intelligence Index | 29.0 | 15.0 |
| Coding Index | 22.0 | 14.2 |
| Math Index | — | 66.3 |
| Output speed (tok/s) | 175.2 | 102.1 |
| Blended price ($/1M) | $3.00 | $0.21 |
| Time to first token (s) | 0.47s | 0.98s |
| aime | — | 72.7% |
| aime 25 | — | 66.3% |
| artificial analysis coding index | 22.00 | 14.20 |
| artificial analysis intelligence index | 29.00 | 15.00 |
| artificial analysis math index | — | 66.30 |
| gpqa | 77.6% | 65.9% |
| hle | 24.2% | 6.8% |
| ifbench | 49.3% | 33.1% |
| lcr | 17.3% | 22.7% |
| livecodebench | — | 51.5% |
| math 500 | — | 97.5% |
| mmlu pro | — | 77.7% |
| scicode | 32.8% | 30.4% |
| tau2 | 59.9% | 10.2% |
| terminalbench hard | 16.7% | 6.1% |
Benchmark data from Artificial Analysis.