AGC-Bench - poetmt: leaderboard
The "poetmt" 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 March 2027. 82 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.6 | 1.55 | #63 |
| 2 | Claude Opus 4.7 | 1.52 | #49 |
| 3 | GPT-5.5 | 1.42 | #31 |
| 4 | GPT-5.4 | 1.41 | #79 |
| 5 | Claude Opus 4.5 | 1.32 | #83 |
| 6 | DeepSeek V3.2 Exp | 1.26 | #234 |
| 7 | DeepSeek V3.1 NEX N1 | 1.25 | #327 |
| 8 | Kimi K2 0905 | 1.2 | #288 |
| 9 | Claude Sonnet 4.5 | 1.2 | #142 |
| 10 | DeepSeek V3.2 | 1.17 | #205 |
| 11 | DeepSeek V3 (0324) | 1.12 | #340 |
| 12 | GLM-5-Turbo | 1.07 | #119 |
| 13 | GLM-5.1 | 1.06 | #97 |
| 14 | GLM-5 | 1.04 | #141 |
| 15 | GPT-5.1 | 1.04 | #135 |
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-poetmt · How It Works · Data refreshed daily, snapshot 2026-10-20.