AGC-Bench - chinese_homophonic_puns: leaderboard
The "chinese_homophonic_puns" 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: Around June 2027. 82 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemma 4 31B (IT) | 1.56 | #176 |
| 2 | Kimi K2.5 | 1.38 | #144 |
| 3 | ERNIE 4.5 300B A47B | 1.38 | #458 |
| 4 | GLM-4.5V | 1.33 | #346 |
| 5 | Claude Opus 4.7 | 1.32 | #49 |
| 6 | GLM-4.7 | 1.28 | #192 |
| 7 | Claude Opus 4.5 | 1.2 | #83 |
| 8 | DeepSeek V3.2 Exp | 1.19 | #234 |
| 9 | Qwen 3.5 Flash (02-23) | 1.18 | #185 |
| 10 | Claude Haiku 4.5 | 1.1 | #280 |
| 11 | DeepSeek V3.1 | 1.02 | #264 |
| 12 | Kimi K2.6 | 1.01 | #104 |
| 13 | Qwen Turbo | 0.99 | #699 |
| 14 | Gemma 4 26B A4B (IT) | 0.95 | #297 |
| 15 | GPT-4.1 Mini | 0.92 | #355 |
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-chinese-homophonic-puns · How It Works · Data refreshed daily, snapshot 2026-10-20.