GPT-6 Astra scores 70.5 on the Respan Index (#2), ahead of Grok 4.5 at 43.2 (#38). GPT-6 Astra leads on 25 of the 25 benchmarks both report. Grok 4.5 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
| 93.7% |
| 57.2% |
| Furniture Assembly (Epoch AI) | 80% | 22.5% |
| GPQA Diamond | 96% | 93.4% |
| LiveBench Agentic Coding (LiveBench) | 57.3% | 56.5% |
| LiveBench Coding (LiveBench) | 80.4% | 68.6% |
| LiveBench Data Analysis (LiveBench) | 83% | 73% |
| LiveBench Instruction Following (LiveBench) | 75.6% | 71.5% |
| LiveBench Language (LiveBench) | 89.4% | 82.8% |
| LiveBench Mathematics (LiveBench) | 96.8% | 90.8% |
| LiveBench Reasoning (LiveBench) | 92.7% | 87.2% |
| OTIS Mock AIME 2024-2025 (Epoch AI) | 100% | 97.8% |
| SAGE (Vals AI) | 46.4% | 35% |
| SimpleQA Verified | 75.6% | 48.3% |
| Terminal-Bench 4.0 (Vals AI) | 59.6% | 8.6% |
| Vending-Bench 2 (Andon Labs) | 15514.7 | 3887.43 |
| WeirdML (Håvard Tveit Ihle) | 93.6% | 46.4% |
| LMArena Agent (LMArena) | 0.1227 | 0.0121 |
| LMArena Elo | 1477 | 1450 |
| LMArena Vision (LMArena) | 1284 | 1279 |
| LMArena WebDev (LMArena) | 1788 | 1552 |
| SimpleBench (SimpleBench) | 86.5% | 70% |
Higher is better. Each score links to where it was published.