Grok 4.6 scores 51.0 on the Respan Index (#19), ahead of Kimi K2.6 at 34.4 (#55). Grok 4.6 leads on 18 of the 22 benchmarks both report. Kimi K2.6 is 1.8x cheaper per token at list price. Grok 4.6 has the larger context window (500K tokens).
* Estimated: no published scores in that category. How the Respan Index works
| 40% |
| 21.7% |
| GPQA Diamond | 94% | 90.8% |
| LiveBench Agentic Coding (LiveBench) | 57% | 46.9% |
| LiveBench Coding (LiveBench) | 76.8% | 78.6% |
| LiveBench Data Analysis (LiveBench) | 73.9% | 65.1% |
| LiveBench Instruction Following (LiveBench) | 71.9% | 64.4% |
| LiveBench Language (LiveBench) | 83.7% | 75.1% |
| LiveBench Mathematics (LiveBench) | 92.6% | 84.3% |
| LiveBench Reasoning (LiveBench) | 90.5% | 79.4% |
| Mystery Game Puzzles (Epoch AI) | 34% | 18% |
| OTIS Mock AIME 2024-2025 (Epoch AI) | 99.2% | 96.1% |
| SAGE (Vals AI) | 28.9% | 50.2% |
| SimpleQA Verified | 49.3% | 34.9% |
| Vending-Bench 2 (Andon Labs) | 9047.03 | 6204.57 |
| WeirdML (Håvard Tveit Ihle) | 67.3% | 55.9% |
| LMArena Elo | 1454 | 1455 |
| LMArena Vision (LMArena) | 1264 | 1265 |
| LMArena WebDev (LMArena) | 1620 | 1509 |
Higher is better. Each score links to where it was published.