Head to head
Z.ai: GLM 5.2 vs Google: Gemini 3.8 Flash
Live prices, context windows, capabilities, and measured quality — updated continuously.
Side by side
| Metric | Z.ai: GLM 5.2 Z Ai | Google: Gemini 3.8 Flash Google Ai Studio |
|---|---|---|
| Input $/M tokens | $0.966 | $0.375 |
| Output $/M tokens | $3.04 | $1.88 |
| Context window | 1.0M | 1.0M |
| Max output | 131K | 66K |
| WM Score | 82 | 82 |
| Benchmark quality | 77.7 | 81.7 |
| Hosts serving it | 25 | 2 |
| Throughput (tok/s) | 156 | 119 |
| Latency (ms) | 295 | 2469 |
| Vision | ||
| Tools | ||
| Reasoning | ||
| JSON mode |
Which should you pick?
Google: Gemini 3.8 Flash costs less per million input tokens — about 2.6× cheaper than Z.ai: GLM 5.2. Google: Gemini 3.8 Flash scores higher on measured quality. 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.