AGC-Bench - metaphor_generation: leaderboard
The "metaphor_generation" 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 July 2027. 83 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.5 | 2.4 | #83 |
| 2 | Claude Opus 4.6 | 2.11 | #62 |
| 3 | GPT-5.5 | 1.33 | #30 |
| 4 | GPT-5.1 | 1.3 | #135 |
| 5 | GPT-5.4 | 1.2 | #80 |
| 6 | Claude Sonnet 4.5 | 1.13 | #141 |
| 7 | jamba-large-1.7 | 1.05 | #525 |
| 8 | Gemma 4 31B (IT) | 1.03 | #176 |
| 9 | GPT-4.1 | 0.97 | #248 |
| 10 | GLM-5-Turbo | 0.93 | #120 |
| 11 | Claude Sonnet 4 | 0.89 | #199 |
| 12 | Gemma 4 26B A4B (IT) | 0.82 | #295 |
| 13 | GPT-4o | 0.79 | #343 |
| 14 | Nova Micro (v1) | 0.78 | #830 |
| 15 | Hermes 3 - Llama-3.1 70B | 0.77 | #926 |
No result here: #4 Claude Opus 5.5, #6 GPT-6 Astra, #7 Gemini 4 Argon.
Interactive version: theaggregate.ai/benchmark?slug=agc-bench-metaphor-generation · How It Works · Data refreshed daily, snapshot 2026-10-21.