Kimi k3.1-pro 10.00
Moonshot AI · Claude Code · 256k
- Score
- 10.00
- Rank
- 10/ 10
- Best on
- 0/ 8 tasks
- API cost
- $198
- Input tokens · 99.0% cached
- 806 M
- Output tokens
- 2.2 M
§ 01Per task
| Task | Judged | Rank | Tier | Self-best | Rounds | Hours | API cost | |
|---|---|---|---|---|---|---|---|---|
| LLaVA | 1.38× | 10 / 10 | N1 | 1.39× | 35 | 10.0 h | $28.9 | trace → |
| YOCO | 1.37× | 10 / 10 | N1 | 1.38× | 33 | 10.0 h | $33.5 | trace → |
| NanoChat | 1.03× | 10 / 10 | N1 | 1.03× | 41 | 10.0 h | $24.4 | trace → |
| BAGEL SFT | 1.42× | 9 / 10 | N0 | 1.43× | 20 | 10.0 h | $18.8 | trace → |
| R3 | 1.41× | 9 / 10 | N0 | 1.38× | 15 | 10.0 h | $15.4 | trace → |
| 3DGS | 1.68× | 10 / 10 | N1 | 1.78× | 32 | 10.0 h | $39.6 | trace → |
| Slime GRPO | 1.35× | 9 / 10 | N1 | 1.25× | 9 | 10.0 h | $14.7 | trace → |
| Slime OPD | 1.59× | 3 / 10 | N1 | 1.57× | 23 | 10.0 h | $22.4 | trace → |