Chinese Classical Bench - Punctuate Punct F1: leaderboard
The "Punctuate Punct F1" subtask of Chinese Classical Bench. A task from Chinese Classical Bench, evaluating comprehension, translation and reasoning over Classical (literary) Chinese texts.
Metric: Score (%). Source: huggingface.co. Saturation forecast: Around January 2027. 10 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.7 | 80.02 | #49 |
| 2 | GLM-5 | 79.88 | #141 |
| 3 | Claude Opus 4.7 (Thinking) | 78.99 | #49 (Claude Opus 4.7) |
| 4 | Claude Sonnet 4.6 | 78.53 | #90 |
| 5 | Qwen3 Coder Next | 76.69 | #326 |
| 6 | Qwen 3.5 35B A3B | 75.31 | #256 |
| 7 | DeepSeek V3.2 | 74.46 | #205 |
| 8 | Claude Haiku 4.5 (20251001) | 72.87 | #261 |
| 9 | MiniMax-M2.5 | 70.86 | #303 |
| 10 | MiniMax-M2.1 | 70.85 | #289 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=chinese-classical-bench-punctuate-punct-f1 · How It Works · Data refreshed daily, snapshot 2026-10-20.