all the models — AI benchmark observatory
← Benchmarks

MEWC

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.

id mewc · max 1 · 1 models reported

#ModelScore
1MiniMax M2.5
MiniMax · open
0.74