MEWC

reasoning

Multilingual: no

MEWC is a benchmark that evaluates AI model performance on multi-environment web challenges, testing agents' ability to navigate and complete complex tasks across diverse web environments.

Methodology

Imported from llm-stats public benchmark metadata. Modality: text. Max score: 1. Categories: agents, reasoning. Language: en. Verified by llm-stats: no.