Qwen3.8-Max scores 50.3 on the Respan Index (#22), ahead of Kimi K2.6 at 34.4 (#55). Qwen3.8-Max leads on 17 of the 19 benchmarks both report. Kimi K2.6 is 1.8x cheaper per token at list price. Qwen3.8-Max has the larger context window (1M tokens).
* Estimated: no published scores in that category. How the Respan Index works
| GPQA Diamond |
| 90.8% |
| 92.6% |
| LiveBench Agentic Coding (LiveBench) | 46.9% | 64.6% |
| LiveBench Coding (LiveBench) | 78.6% | 72.9% |
| LiveBench Data Analysis (LiveBench) | 65.1% | 78.4% |
| LiveBench Instruction Following (LiveBench) | 64.4% | 74.1% |
| LiveBench Language (LiveBench) | 75.1% | 79.7% |
| LiveBench Mathematics (LiveBench) | 84.3% | 91.3% |
| LiveBench Reasoning (LiveBench) | 79.4% | 88.2% |
| Mystery Game Puzzles (Epoch AI) | 18% | 38% |
| OTIS Mock AIME 2024-2025 (Epoch AI) | 96.1% | 99.4% |
| SAGE (Vals AI) | 50.2% | 51.3% |
| SimpleQA Verified | 34.9% | 45.8% |
| LMArena Elo | 1455 | 1481 |
| LMArena Vision (LMArena) | 1265 | 1301 |
| LMArena WebDev (LMArena) | 1509 | 1671 |
Higher is better. Each score links to where it was published.