GPT-5.2 (xhigh) vs Grok 4.20 0309 (Reasoning)
For product management work, compared on the PM Index.
- GPT-5.2 (xhigh) leads on the PM Index (52.8 vs 48.4).
- Grok 4.20 0309 (Reasoning) is cheaper ($3.00 vs $4.81 per 1M tokens).
- Grok 4.20 0309 (Reasoning) is faster (197 vs 78 tok/s).
| Metric | GPT-5.2 (xhigh) | Grok 4.20 0309 (Reasoning) |
|---|---|---|
| PM Index | 52.8 | 48.4 |
| AA Intelligence | 51.3 | 48.5 |
| AA Coding | 48.7 | 42.2 |
| AA Agentic | 60.2 | 50.9 |
| Blended $/1M | $4.81 | $3.00 |
| Output tok/s | 78 | 197 |