AGC-Bench - cue_word_story: leaderboard
The "cue_word_story" 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 April 2027. 83 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GLM-5.1 | 1.25 | #97 |
| 2 | Kimi K2 0905 | 1.16 | #288 |
| 3 | GLM-5-Turbo | 1.15 | #119 |
| 4 | Gemini 2.5 Flash | 1.04 | #245 |
| 5 | GPT-5.5 | 1.04 | #31 |
| 6 | GPT-5.1 | 1.02 | #135 |
| 7 | GPT-5.4 | 1.02 | #79 |
| 8 | Qwen 3.5 Flash (02-23) | 0.96 | #185 |
| 9 | ERNIE-4.5-VL-424B-A47B | 0.96 | #391 |
| 10 | DeepSeek V3.1 | 0.93 | #264 |
| 11 | Claude Sonnet 4.5 | 0.92 | #142 |
| 12 | DeepSeek V3.2 Exp | 0.91 | #234 |
| 13 | Mistral Medium 3.1 | 0.91 | #408 |
| 14 | GLM-4.6 | 0.89 | #252 |
| 15 | ERNIE 4.5 300B A47B | 0.89 | #458 |
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-cue-word-story · How It Works · Data refreshed daily, snapshot 2026-10-20.