Grok 4.3 scores 20.7 on the Respan Index (#92), ahead of GPT-5.4 nano at 20.5 (#93). GPT-5.4 nano leads on 10 of the 17 benchmarks both report. GPT-5.4 nano is 3.4x cheaper per token at list price. Grok 4.3 has the larger context window (1M tokens).
* Estimated: no published scores in that category. How the Respan Index works
| 46.8% |
| 18.5% |
| LiveBench Coding (LiveBench) | 70.8% | 69.9% |
| LiveBench Data Analysis (LiveBench) | 67.6% | 55.8% |
| LiveBench Instruction Following (LiveBench) | 67.2% | 62.8% |
| LiveBench Language (LiveBench) | 62.5% | 73.6% |
| LiveBench Mathematics (LiveBench) | 91% | 84.3% |
| LiveBench Reasoning (LiveBench) | 81.1% | 70.8% |
| OTIS Mock AIME 2024-2025 (Epoch AI) | 87.8% | 93.3% |
| SAGE (Vals AI) | 38.1% | 19.7% |
| SimpleQA Verified | 11.7% | 33.2% |
| WeirdML (Håvard Tveit Ihle) | 49.2% | 49.9% |
| LMArena Elo | 1401 | 1398 |
| LMArena Vision (LMArena) | 1199 | 1243 |
Higher is better. Each score links to where it was published.