← All comparisons
GLM-4.5V (Reasoning) vs Qwen3 235B A22B 2507 Instruct
Z AI vs Alibaba — side-by-side benchmark comparison
| GLM-4.5V (Reasoning) | Qwen3 235B A22B 2507 Instruct | |
|---|---|---|
| Intelligence Index | 15.1 | 25.0 |
| Coding Index | 10.9 | 22.1 |
| Math Index | 73.0 | 71.7 |
| Output speed (tok/s) | 21.3 | 57.0 |
| Blended price ($/1M) | $0.90 | $0.36 |
| Time to first token (s) | 1.13s | 1.34s |
| aime | — | 71.7% |
| aime 25 | 73.0% | 71.7% |
| artificial analysis coding index | 10.90 | 22.10 |
| artificial analysis intelligence index | 15.10 | 25.00 |
| artificial analysis math index | 73.00 | 71.70 |
| gpqa | 68.4% | 75.3% |
| hle | 5.9% | 10.6% |
| ifbench | 34.2% | 46.1% |
| lcr | 0.0% | 31.2% |
| livecodebench | 60.4% | 52.4% |
| math 500 | — | 98.0% |
| mmlu pro | 78.8% | 82.8% |
| scicode | 22.1% | 36.0% |
| tau2 | 22.5% | 33.3% |
| terminalbench hard | 5.3% | 15.2% |
Benchmark data from Artificial Analysis.