{"benchmark_id":"physicsfinals","name":"PhysicsFinals","parent_benchmark":null,"categories":["math","physics","reasoning"],"modality":"text","multilingual":false,"max_score":1.0,"language":"en","description":"PHYSICS is a comprehensive benchmark for university-level physics problem solving, containing 1,297 expert-annotated problems covering six core areas: classical mechanics, quantum mechanics, thermodynamics and statistical mechanics, electromagnetism, atomic physics, and optics. Each problem requires advanced physics knowledge and mathematical reasoning. Even advanced models like o3-mini achieve only 59.9% accuracy.","paper_link":"https://arxiv.org/abs/2503.21821","implementation_link":null,"verified":false,"created_at":"2026-05-07T16:53:25.762699+00:00","updated_at":"2026-07-05T18:27:58.630755+00:00","statistics":{"total_models":2,"average_score":0.6065,"min_score":0.574,"max_score":0.639,"score_stddev":0.04596194077712555,"verified_count":0,"self_reported_count":2},"child_benchmarks":[],"linked_dataset":null,"models":[{"rank":1,"model_id":"gemini-1.5-pro","model_name":"Gemini 1.5 Pro","organization_id":"google","organization_name":"Google","organization_country":"US","score":0.639,"normalized_score":0.639,"verified":false,"self_reported":true,"self_reported_source":"https://arxiv.org/pdf/2403.05530","analysis_method":"0-shot","verification_date":null,"provider_id":null,"input_cost_per_million":null,"output_cost_per_million":null,"context_window":null,"announcement_date":"2024-05-01","param_count":null,"is_open_source":false,"is_new":false,"best_latency":null,"latency_provider":null,"best_throughput":null,"throughput_provider":null,"context_provider":null},{"rank":2,"model_id":"gemini-1.5-flash","model_name":"Gemini 1.5 Flash","organization_id":"google","organization_name":"Google","organization_country":"US","score":0.574,"normalized_score":0.574,"verified":false,"self_reported":true,"self_reported_source":"https://arxiv.org/abs/2303.16416","analysis_method":"Accuracy (0-shot)","verification_date":null,"provider_id":null,"input_cost_per_million":null,"output_cost_per_million":null,"context_window":null,"announcement_date":"2024-05-01","param_count":null,"is_open_source":false,"is_new":false,"best_latency":null,"latency_provider":null,"best_throughput":null,"throughput_provider":null,"context_provider":null}]}