Explicit Edit Benchmark: leaderboard
Metric: Explicit Edit Score v2 (%). Source: huggingface.co. 13 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | DeepSeek V4.1 Flash | 99.3 |
| 2 | GPT-5.6 Sol | 98.9 |
| 3 | DeepSeek V4 Flash | 98.1 |
| 4 | Kimi K2.7 Code | 96.5 |
| 5 | Grok 4.6 | 95.8 |
| 6 | GLM-5.3 Flash | 95.6 |
| 7 | GPT-5.6 Terra | 95 |
| 8 | GPT-5.6 Luna | 94.1 |
| 9 | Qwen 3.7 Plus | 92.4 |
| 10 | Qwen 3.8 Flash | 88.1 |
| 11 | MiMo-V2.5 | 86.5 |
| 12 | MiniMax-M3 | 80.3 |
Interactive version: theaggregate.ai/benchmark?slug=explicit-edit-benchmark · How It Works · Data refreshed daily, snapshot 2026-09-21.