Neo-Classic - Line Matching: leaderboard

Metric: Accuracy (%; select the matching line of a couplet drawn from a contemporary quatrain or regulated verse, four options; accuracy on a 100-question sample of the task, as reported in the paper's full results table (accuracy@IF); temperature 0.1, top-p 0.9). Source: arxiv.org. Saturation forecast: Around 2030. 8 models tracked.

Top models

#ModelScore
1Qwen 3 Max67
2DeepSeek V3.264
3Gemini 3 Pro (Preview)57
4GPT-4o56.6
5Yi 1.5 34B Chat54
6Gemini 2.5 Pro52

Interactive version: theaggregate.ai/benchmark?slug=neo-classic-line-matching · How It Works · Data refreshed daily, snapshot 2026-09-26.