WebForge-Bench - Information Retrieval: leaderboard

WebForge-Bench (2026) tests browser agents on tasks set in automatically generated, reproducible websites with injected real-web noise, controlled for difficulty along seven dimensions. This board is task accuracy in percent on the information retrieval and analysis domain with screenshot and DOM input; higher is better.

Metric: Accuracy (%) on the information retrieval and analysis tasks of the 934 WebForge-Bench tasks on generated, self-contained websites with injected real-web noise, up to 50 browser actions in a Chromium GUI, final-state answer checking; multimodal setting: screenshot plus DOM; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 10 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Gemini 3 Pro79.4#145
2Kimi K2.575.6#266
3Claude Sonnet 4.573.8#261
4GPT-5 Mini73.8#364
5GPT-5.264.4#185
6Gemini 3 Flash62.5#175
7Qwen 3 VL 235B A22B58.8#485
8GPT-5 Nano43.8#906
9Gemini 2.5 Flash Lite41.9#964
10qwen3-omni-30B26.2#1608

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=webforge-bench-information-retrieval · How It Works · Data refreshed daily, snapshot 2026-11-21.