Grok 4.6 scores 51.0 on the Respan Index (#19), ahead of DeepSeek V4 Flash at 43.3 (#37). Grok 4.6 leads on 17 of the 20 benchmarks both report. DeepSeek V4 Flash has the larger context window (1M tokens).
* Estimated: no published scores in that category. How the Respan Index works
| GPQA Diamond |
| 91% |
| 94% |
| LiveBench Agentic Coding (LiveBench) | 46.8% | 57% |
| LiveBench Coding (LiveBench) | 75% | 76.8% |
| LiveBench Data Analysis (LiveBench) | 79.3% | 73.9% |
| LiveBench Instruction Following (LiveBench) | 65.5% | 71.9% |
| LiveBench Language (LiveBench) | 79.2% | 83.7% |
| LiveBench Mathematics (LiveBench) | 86.8% | 92.6% |
| LiveBench Reasoning (LiveBench) | 86.6% | 90.5% |
| Mystery Game Puzzles (Epoch AI) | 34% | 34% |
| OTIS Mock AIME 2024-2025 (Epoch AI) | 94.4% | 99.2% |
| SimpleQA Verified | 33.6% | 49.3% |
| Terminal-Bench 4.0 (Vals AI) | 18.7% | 17.2% |
| WeirdML (Håvard Tveit Ihle) | 63% | 67.3% |
| LMArena Elo | 1436 | 1454 |
| LMArena WebDev (LMArena) | 1581 | 1620 |
| SimpleBench (SimpleBench) | 61.1% | 75.9% |
Higher is better. Each score links to where it was published.