GRACE - 2WikiMHQA Categorization: leaderboard
Metric: Four-class error-categorization macro-F1 (%) on 2WikiMHQA grounding traces; higher is better. Source: arxiv.org. Saturation forecast: Around December 2026. 10 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 3.5 27B | 76.68 |
| 2 | Gemini 3 Flash | 74.86 |
| 3 | GPT-5.4 | 72.81 |
| 4 | Gemini 3.1 Pro (Preview) | 72.53 |
| 5 | Gemma 4 31B | 62.38 |
| 6 | Qwen 3.5 35B A3B | 60.53 |
| 7 | Qwen 3 14B | 50.32 |
| 8 | GPT-5.4 Mini | 49.49 |
| 9 | Qwen 3 8B | 38.89 |
| 10 | Llama 3.1 8B | 28.98 |
Interactive version: theaggregate.ai/benchmark?slug=grace-2wikimhqa-categorization · How It Works · Data refreshed daily, snapshot 2026-09-29.