AGC-Bench - tinystories — leaderboard
Metric: Dataset z-score. Source: huggingface.co. 83 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 3 Next 80B A3B Instruct | 1.8 |
| 2 | Kimi K2 0905 | 1.67 |
| 3 | DeepSeek V3.2 Exp | 1.64 |
| 4 | Mistral Medium 3.1 | 1.6 |
| 5 | Ministral 3 8B | 1.56 |
| 6 | DeepSeek V3.1 Terminus | 1.23 |
| 7 | Gemini 2.5 Flash | 1.2 |
| 8 | GPT-5.1 | 1.19 |
| 9 | Gemma 3 27B (IT) | 1.09 |
| 10 | Kimi K2.6 | 1.02 |
| 11 | Mixtral 8x22B Instruct | 0.94 |
| 12 | GPT-5.4 Nano | 0.91 |
| 13 | DeepSeek V3.1 | 0.88 |
| 14 | nemotron-3-nano-30B-a3B | 0.81 |
| 15 | Kimi K2.5 | 0.81 |
Interactive version: theaggregate.ai/benchmark?slug=agc-bench-tinystories · How the rankings work · Data refreshed daily, snapshot 2026-07-22.