DeepSeek V4 Flash scores 43.3 on the Respan Index (#37), ahead of Grok 4.5 at 43.2 (#38). Grok 4.5 leads on 12 of the 19 benchmarks both report. DeepSeek V4 Flash has the larger context window (1M tokens).
* Estimated: no published scores in that category. How the Respan Index works
| 91% |
| 93.4% |
| LiveBench Agentic Coding (LiveBench) | 46.8% | 56.5% |
| LiveBench Coding (LiveBench) | 75% | 68.6% |
| LiveBench Data Analysis (LiveBench) | 79.3% | 73% |
| LiveBench Instruction Following (LiveBench) | 65.5% | 71.5% |
| LiveBench Language (LiveBench) | 79.2% | 82.8% |
| LiveBench Mathematics (LiveBench) | 86.8% | 90.8% |
| LiveBench Reasoning (LiveBench) | 86.6% | 87.2% |
| OTIS Mock AIME 2024-2025 (Epoch AI) | 94.4% | 97.8% |
| SimpleQA Verified | 33.6% | 48.3% |
| Terminal-Bench 4.0 (Vals AI) | 18.7% | 8.6% |
| WeirdML (Håvard Tveit Ihle) | 63% | 46.4% |
| LMArena Elo | 1436 | 1450 |
| LMArena WebDev (LMArena) | 1581 | 1552 |
| SimpleBench (SimpleBench) | 61.1% | 70% |
Higher is better. Each score links to where it was published.