WebMainBench: leaderboard

Human-annotated web main-content extraction benchmark evaluating full-page extraction systems with ROUGE-style metrics.

Metric: ROUGE-N F1 (N=5, %). Source: huggingface.co. Saturation forecast: Around January 2027. 4 models tracked.

Top models

#ModelScoreOverall rank
1GPT-590.24#98
2Gemini 2.5 Pro89.79#152

No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.

Interactive version: theaggregate.ai/benchmark?slug=webmainbench · How It Works · Data refreshed daily, snapshot 2026-10-20.