Head to head
Z.ai: GLM 5.2 vs Anthropic: Claude Sonnet 4.6
Live prices, context windows, capabilities, and measured quality — updated continuously.
Side by side
| Metric | Z.ai: GLM 5.2 Z Ai | Anthropic: Claude Sonnet 4.6 Anthropic |
|---|---|---|
| Input $/M tokens | $0.966 | $3.00 |
| Output $/M tokens | $3.04 | $15.00 |
| Context window | 1.0M | 1.0M |
| Max output | 131K | 128K |
| WM Score | 85 | 81 |
| Benchmark quality | 82.5 | 76.5 |
| Hosts serving it | 27 | 5 |
| Throughput (tok/s) | 158 | 62 |
| Latency (ms) | 441 | 1051 |
| Vision | ||
| Tools | ||
| Reasoning | ||
| JSON mode |
Which should you pick?
Z.ai: GLM 5.2 costs less per million input tokens — about 3.1× cheaper than Anthropic: Claude Sonnet 4.6. Z.ai: GLM 5.2 scores higher on measured quality. Z.ai: GLM 5.2 takes a larger context window, which matters for long documents or big codebases. Prices here are live list prices; the same model can cost less on a different host, so check the model page for the cheapest offering.