Grok 4.6 vs GLM-5.3
Grok 4.6 (xAI) and GLM-5.3 (Zhipu AI (Z.ai)) compared on price, context window, speed and independent benchmarks. Grok 4.6 scores higher on the Artificial Analysis Intelligence Index (60.9 vs 59.5), while GLM-5.3 is 1.4× cheaper on a blended basis. Pick Grok 4.6 for the hardest tasks and GLM-5.3 when cost dominates.
| Grok 4.6 | GLM-5.3 | |
|---|---|---|
| Provider | xAI | Zhipu AI (Z.ai) |
| Input $/1M | $2.00 | $1.40 |
| Output $/1M | $6.00 | $4.40 |
| Blended $/1M (3:1) | $3.00 | $2.15 |
| Cached input $/1M | $0.50 | $0.26 |
| Context window | 500K | 1M |
| Knowledge cutoff | 2026-02 | — |
| License | proprietary | open weights |
| Speed (tok/s) | 60 | 78 |
| Artificial Analysis Intelligence Index | 60.9 | 59.5 |
| LMArena Elo | 1444 | 1474 |
| GPQA Diamond | 94.9% | 91.7% |
| Humanity's Last Exam | 42.9% | 42.3% |