Grok 4.6 scores 51.0 on the Respan Index (#19), ahead of Kimi K3 at 50.6 (#21). Kimi K3 leads on 16 of the 29 benchmarks both report. Grok 4.6 is 2.0x cheaper per token at list price. Kimi K3 has the larger context window (1.0M tokens).
* Estimated: no published scores in that category. How the Respan Index works
| 66% |
| 72.2% |
| FrontierSWE V2 (Proximal Labs) | 25.3% | 25.9% |
| Furniture Assembly (Epoch AI) | 40% | 34.2% |
| GPQA Diamond | 94% | 93.5% |
| LiveBench Agentic Coding (LiveBench) | 57% | 62.2% |
| LiveBench Coding (LiveBench) | 76.8% | 81.4% |
| LiveBench Data Analysis (LiveBench) | 73.9% | 78.7% |
| LiveBench Instruction Following (LiveBench) | 71.9% | 71.4% |
| LiveBench Language (LiveBench) | 83.7% | 85.5% |
| LiveBench Mathematics (LiveBench) | 92.6% | 84.4% |
| LiveBench Reasoning (LiveBench) | 90.5% | 90.7% |
| Mystery Game Puzzles (Epoch AI) | 34% | 26% |
| OTIS Mock AIME 2024-2025 (Epoch AI) | 99.2% | 97.2% |
| SAGE (Vals AI) | 28.9% | 52.8% |
| SimpleQA Verified | 49.3% | 50.6% |
| Terminal-Bench 4.0 (Vals AI) | 17.2% | 17.2% |
| Vending-Bench 2 (Andon Labs) | 9047.03 | 5165.04 |
| WeirdML (Håvard Tveit Ihle) | 67.3% | 82.6% |
| LMArena Agent (LMArena) | 0.0128 | 0.0418 |
| LMArena Elo | 1454 | 1488 |
| LMArena WebDev (LMArena) | 1620 | 1658 |
| SimpleBench (SimpleBench) | 75.9% | 60.7% |
| AA-Briefcase | 1577 | 1548 |
| APEX-Agents | 57.5% | 41% |
| DeepSWE v1.1 | 65.9% | 67.3% |
Higher is better. Each score links to where it was published.