{"benchmark_id":"phybench","benchmark_name":"PHYBench","benchmark_description":"PHYBench is a benchmark of real-world physics problems spanning mechanics, electromagnetism, thermodynamics, optics, and modern physics, designed to evaluate physical perception and multi-step quantitative reasoning in large language models.","max_score":1.0,"categories":["physics","reasoning","science"],"modality":"text","total_models":1,"entries":[{"rank":1,"model_id":"hy3","model_name":"Hy3","organization_name":"Tencent","organization_id":"tencent","benchmark_score":0.774,"normalized_score":0.774,"verified":false,"self_reported":true,"provider_id":null,"input_cost_per_million":null,"output_cost_per_million":null,"speed_rps":null,"context_window":null,"release_date":"2026-07-06","announcement_date":"2026-07-06","multimodal":false,"param_count":295000000000,"is_new":false}]}