AGC-Bench - pun_eval — leaderboard
Metric: Dataset z-score. Source: huggingface.co. 83 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | DeepSeek V3 (0324) | 1.52 |
| 2 | Mistral Medium 3.1 | 1.42 |
| 3 | GPT-5.4 Mini | 1.4 |
| 4 | Claude Opus 4.7 | 1.33 |
| 5 | DeepSeek V3 Chat | 1.22 |
| 6 | Ministral 3 8B | 1.22 |
| 7 | GPT-5.5 | 1.04 |
| 8 | Gemma 4 26B A4B (IT) | 0.94 |
| 9 | GPT-5.4 | 0.93 |
| 10 | Claude Opus 4.5 | 0.9 |
| 11 | GPT-5.4 Nano | 0.87 |
| 12 | NVIDIA Nemotron Nano 9B V2 | 0.83 |
| 13 | Qwen 3 Next 80B A3B Instruct | 0.8 |
| 14 | Llama 3 8B Instruct | 0.75 |
| 15 | Kimi K2 0905 | 0.7 |
Interactive version: theaggregate.ai/benchmark?slug=agc-bench-pun-eval · How the rankings work · Data refreshed daily, snapshot 2026-07-22.