DeepSeek: R1 wins on more metrics (5 of 8), but the right pick depends on what you optimize for — see the breakdown below.
| Metric | DeepSeek: R1 | Nous: Hermes 4 405B |
|---|---|---|
| Intelligence Index | 18.8✓ | 18.6 |
| Coding Index | 15.9 | 16.0✓ |
| GPQA Diamond | 71% | 73%✓ |
| Design Arena Elo | — | — |
| Speed (tokens/sec) | 72✓ | 31 |
| Latency | 1.4s | 353ms✓ |
| Input price /M | $0.700✓ | $1.00 |
| Output price /M | $2.50✓ | $3.00 |
| Context window | 164K✓ | 131K |
| Capabilities | ReasoningToolsJSON | ReasoningJSON |