AGC-Bench - future_ideas: leaderboard
The "future_ideas" 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 April 2027. 83 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Kimi K2 0905 | 2.07 | #288 |
| 2 | Qwen 3 Next 80B A3B Instruct | 1.62 | #368 |
| 3 | Claude Opus 4.6 | 1.45 | #63 |
| 4 | Kimi K2.6 | 1.38 | #104 |
| 5 | Mistral Medium 3.1 | 1.32 | #408 |
| 6 | GLM-5-Turbo | 1.29 | #119 |
| 7 | Kimi K2.5 | 1.29 | #144 |
| 8 | Claude Sonnet 4.5 | 1.21 | #142 |
| 9 | Nemotron 3 Super | 1.18 | #335 |
| 10 | Claude Sonnet 4.6 | 1.15 | #90 |
| 11 | GLM-5.1 | 1.08 | #97 |
| 12 | Mistral Large 3 | 1.07 | #401 |
| 13 | GPT-5.4 | 1.06 | #79 |
| 14 | Claude Sonnet 4 | 1.01 | #199 |
| 15 | Ministral 3 8B | 0.95 | #687 |
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-future-ideas · How It Works · Data refreshed daily, snapshot 2026-10-20.