xAI: Grok 4.20 wins on more metrics (5 of 9), but the right pick depends on what you optimize for — see the breakdown below.
| Metric | DeepSeek: DeepSeek V3.1 Terminus | xAI: Grok 4.20 |
|---|---|---|
| Intelligence Index | 28.5 | 29.7✓ |
| Coding Index | 31.9✓ | 25.4 |
| GPQA Diamond | 75% | 79%✓ |
| Design Arena Elo | 1238✓ | — |
| Speed (tokens/sec) | 28 | 76✓ |
| Latency | 848ms | 707ms✓ |
| Input price /M | $0.270✓ | $1.25 |
| Output price /M | $0.950✓ | $2.50 |
| Context window | 164K | 2M✓ |
| Capabilities | ReasoningToolsJSON | ReasoningToolsJSONVision |