Grok 4.5 scores 43.2 on the Respan Index (#38), ahead of Gemini 3.5 Flash at 42.3 (#40). Gemini 3.5 Flash leads on 13 of the 23 benchmarks both report. Grok 4.5 is 1.1x cheaper per token at list price. Gemini 3.5 Flash has the larger context window (1.0M tokens).
* Estimated: no published scores in that category. How the Respan Index works
| 62.8% |
| 57.2% |
| GPQA Diamond | 92.8% | 93.4% |
| LiveBench Agentic Coding (LiveBench) | 49% | 56.5% |
| LiveBench Coding (LiveBench) | 78.2% | 68.6% |
| LiveBench Data Analysis (LiveBench) | 64.9% | 73% |
| LiveBench Instruction Following (LiveBench) | 75.6% | 71.5% |
| LiveBench Language (LiveBench) | 84.6% | 82.8% |
| LiveBench Mathematics (LiveBench) | 88.2% | 90.8% |
| LiveBench Reasoning (LiveBench) | 82% | 87.2% |
| OTIS Mock AIME 2024-2025 (Epoch AI) | 95.6% | 97.8% |
| SAGE (Vals AI) | 49.9% | 35% |
| SimpleQA Verified | 66.2% | 48.3% |
| Terminal-Bench 4.0 (Vals AI) | 6.1% | 8.6% |
| Vending-Bench 2 (Andon Labs) | 5396.42 | 3887.43 |
| WeirdML (Håvard Tveit Ihle) | 62.6% | 46.4% |
| LMArena Elo | 1477 | 1450 |
| LMArena Vision (LMArena) | 1284 | 1279 |
| LMArena WebDev (LMArena) | 1499 | 1552 |
| SimpleBench (SimpleBench) | 76.7% | 70% |
Higher is better. Each score links to where it was published.