Kimi K2.7 Code scores 34.1 on the Respan Index (#58), ahead of Qwen3.6-Plus at 33.2 (#60). They split the 14 benchmarks both report evenly. Qwen3.6-Plus is 1.5x cheaper per token at list price. Qwen3.6-Plus has the larger context window (1M tokens).
* Estimated: no published scores in that category. How the Respan Index works
| 74% |
| 78.2% |
| LiveBench Data Analysis (LiveBench) | 62.7% | 69.9% |
| LiveBench Instruction Following (LiveBench) | 56.3% | 58.3% |
| LiveBench Language (LiveBench) | 77.9% | 75% |
| LiveBench Mathematics (LiveBench) | 79.6% | 83.7% |
| LiveBench Reasoning (LiveBench) | 82.8% | 75.8% |
| OTIS Mock AIME 2024-2025 (Epoch AI) | 95.6% | 93.3% |
| SimpleQA Verified | 36.5% | 44.1% |
| Vending-Bench 2 (Andon Labs) | 5082.94 | 5114.87 |
| LMArena WebDev (LMArena) | 1473 | 1461 |
Higher is better. Each score links to where it was published.