GPT-6 Astra scores 70.5 on the Respan Index (#2), ahead of Grok 4.6 at 51.0 (#19). GPT-6 Astra leads on 27 of the 28 benchmarks both report. Grok 4.6 is 6.7x cheaper per token at list price. GPT-6 Astra has the larger context window (1.1M tokens).
* Estimated: no published scores in that category. How the Respan Index works
| 97.6% |
| 31.7% |
| FrontierMath Tiers 1-3 v2 (Epoch AI) | 93.7% | 66% |
| FrontierSWE V2 (Proximal Labs) | 65.5% | 25.3% |
| Furniture Assembly (Epoch AI) | 80% | 40% |
| GPQA Diamond | 96% | 94% |
| LiveBench Agentic Coding (LiveBench) | 57.3% | 57% |
| LiveBench Coding (LiveBench) | 80.4% | 76.8% |
| LiveBench Data Analysis (LiveBench) | 83% | 73.9% |
| LiveBench Instruction Following (LiveBench) | 75.6% | 71.9% |
| LiveBench Language (LiveBench) | 89.4% | 83.7% |
| LiveBench Mathematics (LiveBench) | 96.8% | 92.6% |
| LiveBench Reasoning (LiveBench) | 92.7% | 90.5% |
| Mystery Game Puzzles (Epoch AI) | 84% | 34% |
| OTIS Mock AIME 2024-2025 (Epoch AI) | 100% | 99.2% |
| SAGE (Vals AI) | 46.4% | 28.9% |
| SimpleQA Verified | 75.6% | 49.3% |
| Terminal-Bench 4.0 (Vals AI) | 59.6% | 17.2% |
| Vending-Bench 2 (Andon Labs) | 15514.7 | 9047.03 |
| WeirdML (Håvard Tveit Ihle) | 93.6% | 67.3% |
| LMArena Agent (LMArena) | 0.1227 | 0.0128 |
| LMArena Elo | 1477 | 1454 |
| LMArena Vision (LMArena) | 1284 | 1264 |
| LMArena WebDev (LMArena) | 1788 | 1620 |
| SimpleBench (SimpleBench) | 86.5% | 75.9% |
Higher is better. Each score links to where it was published.