{"benchmark_id":"critpt","benchmark_name":"CritPT","benchmark_description":"CritPT is a challenging reasoning benchmark reported by Qwen for evaluating frontier mathematical and critical problem-solving capability.","max_score":1.0,"categories":["math","reasoning"],"modality":"text","total_models":5,"entries":[{"rank":1,"model_id":"glm-5.2","model_name":"GLM-5.2","organization_name":"Zhipu AI","organization_id":"zai-org","benchmark_score":0.167,"normalized_score":0.167,"verified":false,"self_reported":true,"provider_id":"deepinfra","input_cost_per_million":0.95,"output_cost_per_million":3.0,"speed_rps":null,"context_window":1048576,"release_date":"2026-06-16","announcement_date":"2026-06-16","multimodal":false,"param_count":753000000000,"is_new":false},{"rank":2,"model_id":"qwen3.7-max","model_name":"Qwen3.7 Max","organization_name":"Alibaba Cloud / Qwen Team","organization_id":"qwen","benchmark_score":0.114,"normalized_score":0.114,"verified":false,"self_reported":true,"provider_id":"novita","input_cost_per_million":1.25,"output_cost_per_million":3.75,"speed_rps":null,"context_window":1000000,"release_date":"2026-05-19","announcement_date":"2026-05-19","multimodal":false,"param_count":null,"is_new":false},{"rank":3,"model_id":"inkling-small","model_name":"Inkling-Small","organization_name":"Thinking Machines Lab","organization_id":"thinking-machines","benchmark_score":0.083,"normalized_score":0.083,"verified":false,"self_reported":true,"provider_id":"thinking-machines","input_cost_per_million":0.3,"output_cost_per_million":1.2,"speed_rps":null,"context_window":256000,"release_date":"2026-07-30","announcement_date":"2026-07-30","multimodal":true,"param_count":276000000000,"is_new":false},{"rank":4,"model_id":"qwen3.7-plus","model_name":"Qwen3.7-Plus","organization_name":"Alibaba Cloud / Qwen Team","organization_id":"qwen","benchmark_score":0.06,"normalized_score":0.06,"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":"2026-05-31","announcement_date":"2026-05-31","multimodal":true,"param_count":null,"is_new":false},{"rank":5,"model_id":"nemotron-3-ultra-550b-a55b","model_name":"Nemotron 3 Ultra (550B A55B)","organization_name":"NVIDIA","organization_id":"nvidia","benchmark_score":0.031,"normalized_score":0.031,"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":"2026-06-04","announcement_date":"2026-06-04","multimodal":false,"param_count":550000000000,"is_new":false}]}