Gemini 3.1 Pro Preview vs GPT-5.2-Codex
| Gemini 3.1 Pro Preview | GPT-5.2-Codex | |
|---|---|---|
| Coding index | 32.3 | 45.4 |
| Benchmarks behind it | Verified, 4 benchmarks | Verified, 3 benchmarks |
| Output $/Mtok | $12.00 | $14.00 |
| Cost per solved task | $0.99 | $0.43 |
| Context window | 1.05M | 400K |