AGC-Bench - lcc_metaphor: leaderboard
The "lcc_metaphor" 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 August 2027. 82 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3.5 Flash (02-23) | 1.34 | #185 |
| 2 | Claude Opus 4.6 | 1.26 | #63 |
| 3 | Claude Opus 4.5 | 1.26 | #83 |
| 4 | GLM-4.7 | 1.21 | #192 |
| 5 | GPT-5.4 | 1.18 | #79 |
| 6 | GPT-4.1 Mini | 1.18 | #355 |
| 7 | Gemini 2.5 Flash | 1.1 | #245 |
| 8 | Claude Opus 4.7 | 1.1 | #49 |
| 9 | GPT-4o Mini | 1.03 | #599 |
| 10 | GPT-5.5 | 1.02 | #31 |
| 11 | GPT-5.4 Mini | 1.02 | #212 |
| 12 | Hermes 4 70B | 1.02 | #506 |
| 13 | ERNIE 4.5 300B A47B | 0.95 | #458 |
| 14 | GPT-5.1 | 0.94 | #135 |
| 15 | GLM-5-Turbo | 0.94 | #119 |
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-lcc-metaphor · How It Works · Data refreshed daily, snapshot 2026-10-20.