← All comparisons
Grok 4.3 (high) vs Qwen2.5 Instruct 72B
xAI vs Alibaba — side-by-side benchmark comparison
| Grok 4.3 (high) | Qwen2.5 Instruct 72B | |
|---|---|---|
| Intelligence Index | 53.2 | 15.6 |
| Coding Index | 41.0 | 11.9 |
| Math Index | — | 14.0 |
| Output speed (tok/s) | 161.0 | 55.4 |
| Blended price ($/1M) | $1.56 | $0.37 |
| Time to first token (s) | 22.30s | 1.23s |
| aime | — | 16.0% |
| aime 25 | — | 14.0% |
| artificial analysis coding index | 41.00 | 11.90 |
| artificial analysis intelligence index | 53.20 | 15.60 |
| artificial analysis math index | — | 14.00 |
| gpqa | 90.1% | 49.1% |
| hle | 35.0% | 4.2% |
| ifbench | 81.3% | 36.9% |
| lcr | 64.3% | 20.3% |
| livecodebench | — | 27.6% |
| math 500 | — | 85.8% |
| mmlu pro | — | 72.0% |
| scicode | 47.3% | 26.7% |
| tau2 | 97.7% | 34.5% |
| terminalbench hard | 37.9% | 4.5% |
Benchmark data from Artificial Analysis.