DeepSeek V3.2 Exp (Non-reasoning) vs Grok 4.1 Fast (Reasoning)
For product management work, compared on the PM Index.
- Grok 4.1 Fast (Reasoning) leads on the PM Index (40.0 vs 29.1).
| Metric | DeepSeek V3.2 Exp (Non-reasoning) | Grok 4.1 Fast (Reasoning) |
|---|---|---|
| PM Index | 29.1 | 40.0 |
| AA Intelligence | 28.4 | 38.6 |
| AA Coding | 30.0 | 30.9 |
| AA Agentic | 31.0 | 49.3 |
| Blended $/1M | $0.31 | — |
| Output tok/s | — | — |