AGC-Bench - schnovel: leaderboard
The "schnovel" 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. Saturation forecast: Not forecast. 83 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GLM-5 | 2.04 | #141 |
| 2 | Llama 3.2 11B Instruct | 1.91 | #1126 |
| 3 | Ministral 3 8B | 1.63 | #687 |
| 4 | GLM-4.7 | 1.56 | #192 |
| 5 | Hermes 4 405B | 1.56 | #334 |
| 6 | GLM-5-Turbo | 1.42 | #119 |
| 7 | GPT-4o Mini | 1.35 | #599 |
| 8 | Qwen Turbo | 1.35 | #699 |
| 9 | GPT-4.1 | 1.2 | #248 |
| 10 | DeepSeek V3.1 NEX N1 | 1.2 | #327 |
| 11 | Gemma 4 31B (IT) | 1.13 | #176 |
| 12 | Qwen 3.5 Flash (02-23) | 1.13 | #185 |
| 13 | GPT-5.4 | 1.06 | #79 |
| 14 | GPT-5.4 Nano | 1.06 | #328 |
| 15 | ERNIE-4.5-VL-424B-A47B | 1.06 | #391 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=agc-bench-schnovel · How It Works · Data refreshed daily, snapshot 2026-10-20.