AGC-Bench - Story / Narrative: leaderboard
The "Story / Narrative" 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: JRT z-score. Source: huggingface.co. Saturation forecast: Around March 2027. 83 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.4 | 0.89 | #79 |
| 2 | GPT-5.5 | 0.75 | #31 |
| 3 | GPT-5.1 | 0.73 | #135 |
| 4 | Mistral Medium 3.1 | 0.73 | #408 |
| 5 | GLM-5.1 | 0.71 | #97 |
| 6 | Kimi K2.5 | 0.7 | #144 |
| 7 | Kimi K2 0905 | 0.68 | #288 |
| 8 | Kimi K2.6 | 0.63 | #104 |
| 9 | Claude Opus 4.7 | 0.62 | #49 |
| 10 | DeepSeek V3.1 Terminus | 0.6 | #220 |
| 11 | GLM-5-Turbo | 0.6 | #119 |
| 12 | DeepSeek V3.2 Exp | 0.59 | #234 |
| 13 | Claude Opus 4.5 | 0.58 | #83 |
| 14 | Claude Sonnet 4.6 | 0.56 | #90 |
| 15 | Gemini 2.5 Flash | 0.55 | #245 |
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-story-narrative · How It Works · Data refreshed daily, snapshot 2026-10-20.