AGC-Bench - historical_analogy: leaderboard
The "historical_analogy" 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: Not forecast. 81 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GLM-4.7 | 1.94 | #192 |
| 2 | GLM-5-Turbo | 1.94 | #119 |
| 3 | Claude Opus 4.7 | 1.84 | #49 |
| 4 | GPT-4o | 1.58 | #343 |
| 5 | Kimi K2.5 | 1.58 | #144 |
| 6 | GLM-5.1 | 1.48 | #97 |
| 7 | Kimi K2.6 | 1.41 | #104 |
| 8 | Hermes 4 70B | 1.41 | #506 |
| 9 | Hermes 4 405B | 1.41 | #334 |
| 10 | DeepSeek V3.2 Exp | 1.38 | #234 |
| 11 | DeepSeek V3.1 | 1.31 | #264 |
| 12 | Gemma 4 31B (IT) | 1.23 | #176 |
| 13 | GPT-4.1 | 1.11 | #248 |
| 14 | Gemma 4 26B A4B (IT) | 1.05 | #297 |
| 15 | Hunyuan A13B-Instruct | 1.05 | #666 |
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-historical-analogy · How It Works · Data refreshed daily, snapshot 2026-10-20.