AGC-Bench - futuregen: leaderboard
The "futuregen" subtask of AGC-Bench. A task from AGC-Bench, evaluating models on open-ended creative and figurative-language generation — humor, metaphor, analogy, storytelling and brainstorming — typically graded by LLM-as-judge.
Metric: Dataset z-score. Source: huggingface.co. 81 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4.1 | 2.15 | #248 |
| 2 | Hunyuan A13B-Instruct | 1.52 | #666 |
| 3 | granite-4.0-h-micro | 1.51 | #1088 |
| 4 | Nova Lite (v1) | 1.35 | #830 |
| 5 | Hermes 4 70B | 1.33 | #506 |
| 6 | Mistral Small 3.1 | 1.28 | #610 |
| 7 | GPT-4o Mini | 1.24 | #599 |
| 8 | trinity-large-preview | 1.2 | #582 |
| 9 | Nova Pro (v1) | 1.18 | #497 |
| 10 | GPT-4o | 1.18 | #343 |
| 11 | Hermes 4 405B | 1.17 | #334 |
| 12 | Nova Micro (v1) | 0.88 | #831 |
| 13 | minimax-m2-her | 0.87 | |
| 14 | GLM-4.7 | 0.84 | #192 |
| 15 | Gemma 4 31B (IT) | 0.81 | #176 |
Interactive version: theaggregate.ai/benchmark?slug=agc-bench-futuregen · How It Works · Data refreshed daily, snapshot 2026-10-20.