DeepSeek V3.2 Exp (Non-reasoning) vs GPT-5.2 (xhigh)
For product management work, compared on the PM Index.
- GPT-5.2 (xhigh) leads on the PM Index (52.8 vs 29.1).
- DeepSeek V3.2 Exp (Non-reasoning) is cheaper ($0.31 vs $4.81 per 1M tokens).
| Metric | DeepSeek V3.2 Exp (Non-reasoning) | GPT-5.2 (xhigh) |
|---|---|---|
| PM Index | 29.1 | 52.8 |
| AA Intelligence | 28.4 | 51.3 |
| AA Coding | 30.0 | 48.7 |
| AA Agentic | 31.0 | 60.2 |
| Blended $/1M | $0.31 | $4.81 |
| Output tok/s | — | 78 |