ExpertLongBench (Checklist F1): leaderboard
ExpertLongBench v1 (arXiv:2506.01241v3): mean CLEAR structured-checklist F1 score (0-100%) across 11 expert-level long-form generation tasks spanning 9 domains (outputs up to >5,000 tokens, greedy decoding temp=0); higher is better.
Metric: Mean Checklist F1 (%). Source: huggingface.co. Saturation forecast: Around 2031. 13 models tracked. Updated 21 Nov 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 2.5 Pro | 33.4 | #276 |
| 2 | GPT-5 | 31 | #178 |
| 3 | Gemini 2.0 Flash | 26.8 | #723 |
| 4 | GPT-4o (2024-11-20) | 26.5 | #859 |
| 5 | GPT-4o Mini (2024-07-18) | 26.1 | #1484 |
| 6 | Llama 3.3 70B Instruct | 24.6 | #1385 |
| 7 | Mistral Large 2 (Nov) Instruct (2411) | 24 | #1217 |
| 8 | Qwen 2.5 72B | 23.8 | #1188 |
| 9 | Mistral Nemo Instruct (2407) | 23.8 | #3118 |
| 10 | Llama 3.1 8B Instruct | 23.4 | #4600 |
| 11 | Claude 3.7 Sonnet | 23.2 | #470 |
| 12 | Qwen 2.5 7B | 20.5 | #2929 |
| 13 | Claude 3.5 Haiku | 19.3 | #1580 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=expertlongbench-checklist-f1 · How It Works · Data refreshed daily, snapshot 2026-11-21.