AGC-Bench - c3_crosstalk: leaderboard
The "c3_crosstalk" 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 September 2027. 83 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.7 | 2.02 | #49 |
| 2 | Claude 3.7 Sonnet | 1.83 | #249 |
| 3 | GLM-4.7 | 1.42 | #192 |
| 4 | Claude Sonnet 4.5 | 1.39 | #142 |
| 5 | Claude Opus 4.5 | 1.35 | #83 |
| 6 | Llama 3.3 70B Instruct | 1.13 | #534 |
| 7 | GPT-4o Mini | 1.09 | #599 |
| 8 | jamba-large-1.7 | 1.07 | #525 |
| 9 | Llama 4 Scout | 1.07 | #656 |
| 10 | GPT-4o | 1.05 | #343 |
| 11 | GLM-5 | 1.02 | #141 |
| 12 | Claude Opus 4.6 | 1 | #63 |
| 13 | Hermes 4 405B | 0.96 | #334 |
| 14 | GLM-5-Turbo | 0.95 | #119 |
| 15 | GPT-4.1 | 0.94 | #248 |
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-c3-crosstalk · How It Works · Data refreshed daily, snapshot 2026-10-20.