AGC-Bench - ss_gen: leaderboard
The "ss_gen" 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 | GPT-5.4 | 1.51 | #79 |
| 2 | Qwen 3.5 Flash (02-23) | 1.38 | #185 |
| 3 | DeepSeek V3.1 Terminus | 1.34 | #220 |
| 4 | GPT-5.5 | 1.29 | #31 |
| 5 | Claude Sonnet 4.6 | 1.2 | #90 |
| 6 | GLM-5.1 | 1.14 | #97 |
| 7 | GLM-5 | 1.13 | #141 |
| 8 | Claude Opus 4.6 | 1.12 | #63 |
| 9 | Gemma 4 26B A4B (IT) | 1.06 | #297 |
| 10 | DeepSeek V3.2 | 1.05 | #205 |
| 11 | GLM-5-Turbo | 1.02 | #119 |
| 12 | DeepSeek V3.1 | 1.01 | #264 |
| 13 | Qwen 3 Next 80B A3B Instruct | 0.96 | #368 |
| 14 | GPT-5.1 | 0.96 | #135 |
| 15 | Nemotron 3 Super | 0.95 | #335 |
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-ss-gen · How It Works · Data refreshed daily, snapshot 2026-10-20.