GPT-6 Sol scores 62.8 on the Respan Index (#8), ahead of GPT-5.6 Sol at 60.6 (#10). They split the 23 benchmarks both report evenly. GPT-6 Sol is 2.0x cheaper per token at list price.
* Estimated: no published scores in that category. How the Respan Index works
| Furniture Assembly (Epoch AI) |
| 56.7% |
| 58.3% |
| GPQA Diamond | 94.6% | 94.3% |
| LiveBench Agentic Coding (LiveBench) | 56.2% | 52.9% |
| LiveBench Coding (LiveBench) | 83.9% | 81.8% |
| LiveBench Data Analysis (LiveBench) | 79.8% | 81.2% |
| LiveBench Instruction Following (LiveBench) | 71.8% | 68.6% |
| LiveBench Language (LiveBench) | 87.7% | 85.3% |
| LiveBench Mathematics (LiveBench) | 96.2% | 96.4% |
| LiveBench Reasoning (LiveBench) | 91.7% | 88.7% |
| Mystery Game Puzzles (Epoch AI) | 58% | 56% |
| OTIS Mock AIME 2024-2025 (Epoch AI) | 100% | 100% |
| SAGE (Vals AI) | 52.6% | 44.8% |
| SimpleQA Verified | 69.7% | 60.7% |
| Terminal-Bench 4.0 (Vals AI) | 37.9% | 44.4% |
| Vending-Bench 2 (Andon Labs) | 9619.37 | 14427.85 |
| LMArena Agent (LMArena) | 0.0645 | 0.0971 |
| LMArena Elo | 1484 | 1457 |
| DeepSWE v1.1 | 72.7% | 68.8% |
| SimpleBench (SimpleBench) | 71.7% | 73.1% |
Higher is better. Each score links to where it was published.