Grok 4.6 scores 51.0 on the Respan Index (#19), ahead of Qwen3.7-Max at 40.5 (#45). Grok 4.6 leads on 15 of the 18 benchmarks both report. Grok 4.6 is 1.3x cheaper per token at list price. Qwen3.7-Max has the larger context window (1M tokens).
* Estimated: no published scores in that category. How the Respan Index works
| 94% |
| 90.9% |
| LiveBench Agentic Coding (LiveBench) | 57% | 43.6% |
| LiveBench Coding (LiveBench) | 76.8% | 74.2% |
| LiveBench Data Analysis (LiveBench) | 73.9% | 71.8% |
| LiveBench Instruction Following (LiveBench) | 71.9% | 74% |
| LiveBench Language (LiveBench) | 83.7% | 79.7% |
| LiveBench Mathematics (LiveBench) | 92.6% | 85.2% |
| LiveBench Reasoning (LiveBench) | 90.5% | 83.3% |
| Mystery Game Puzzles (Epoch AI) | 34% | 32% |
| OTIS Mock AIME 2024-2025 (Epoch AI) | 99.2% | 95.6% |
| SimpleQA Verified | 49.3% | 55.8% |
| LMArena Agent (LMArena) | 0.0128 | -0.0513 |
| LMArena WebDev (LMArena) | 1620 | 1515 |
| SimpleBench (SimpleBench) | 75.9% | 70.4% |
Higher is better. Each score links to where it was published.