GPT-6 Astra scores 70.5 on the Respan Index (#2), ahead of Gemini 3.7 Flash at 48.6 (#26). GPT-6 Astra leads on 18 of the 24 benchmarks both report. Gemini 3.7 Flash is 13x 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
| FrontierMath Tiers 1-3 v2 (Epoch AI) |
| 71.6% |
| 93.7% |
| FrontierSWE V2 (Proximal Labs) | 20.3% | 65.5% |
| Furniture Assembly (Epoch AI) | 26.7% | 80% |
| GPQA Diamond | 94.8% | 96% |
| LiveBench Agentic Coding (LiveBench) | 58.3% | 57.3% |
| LiveBench Coding (LiveBench) | 78.9% | 80.4% |
| LiveBench Data Analysis (LiveBench) | 68% | 83% |
| LiveBench Instruction Following (LiveBench) | 79.9% | 75.6% |
| LiveBench Language (LiveBench) | 85.5% | 89.4% |
| LiveBench Mathematics (LiveBench) | 93.5% | 96.8% |
| LiveBench Reasoning (LiveBench) | 87.8% | 92.7% |
| Mystery Game Puzzles (Epoch AI) | 37% | 84% |
| OTIS Mock AIME 2024-2025 (Epoch AI) | 97.2% | 100% |
| SAGE (Vals AI) | 49.2% | 46.4% |
| SimpleQA Verified | 69.2% | 75.6% |
| Terminal-Bench 4.0 (Vals AI) | 12.1% | 59.6% |
| LMArena Agent (LMArena) | -0.0149 | 0.1227 |
| LMArena Elo | 1488 | 1477 |
| LMArena Vision (LMArena) | 1296 | 1284 |
| LMArena WebDev (LMArena) | 1592 | 1788 |
Higher is better. Each score links to where it was published.