FilBench - Generation: leaderboard
Metric: Category Score (%). Source: huggingface.co. 40 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4o (2024-08-06) | 54.65 |
| 2 | GPT-4o Mini | 47.88 |
| 3 | Llama 4 Maverick Instruct FP8 | 33.67 |
| 4 | DeepSeek V3 | 27.53 |
| 5 | Llama 3.1 70B Instruct | 24.03 |
| 6 | Llama 4 Scout Instruct | 19.75 |
| 7 | gemma2-9B-cpt-sea-lionv3-instruct | 17.1 |
| 8 | Sailor2-20B-Chat | 15.95 |
| 9 | Gemma 2 27B (IT) | 15.38 |
| 10 | Qwen 2.5 72B Instruct | 14.98 |
| 11 | Sailor2-8B-Chat | 14.29 |
| 12 | Qwen 2 72B Instruct | 12.11 |
| 13 | Gemma 2 9B (IT) | 11.51 |
| 14 | Mixtral 8x22B Instruct (v0.1) | 10.7 |
| 15 | Llama 3.1 8B CPT SEA LIONv3 Instruct | 10.6 |
Interactive version: theaggregate.ai/benchmark?slug=filbench-generation · How It Works · Data refreshed daily, snapshot 2026-09-19.