{"benchmark_id":"mewc","benchmark_name":"MEWC","benchmark_description":"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.","max_score":1.0,"categories":["reasoning","agents"],"modality":"text","total_models":1,"entries":[{"rank":1,"model_id":"minimax-m2.5","model_name":"MiniMax M2.5","organization_name":"MiniMax","organization_id":"minimax","benchmark_score":0.744,"normalized_score":0.744,"verified":false,"self_reported":true,"provider_id":"minimax","input_cost_per_million":0.3,"output_cost_per_million":1.2,"speed_rps":100.0,"context_window":1000000,"release_date":"2026-02-12","announcement_date":"2026-02-12","multimodal":false,"param_count":230000000000,"is_new":false}]}