{"benchmark_id":"translation-set1→en-comet22","name":"Translation Set1→en COMET22","parent_benchmark":null,"categories":["language"],"modality":"text","multilingual":true,"max_score":1.0,"language":"en","description":"COMET-22 is a neural machine translation evaluation metric that uses an ensemble of two models: a COMET estimator trained with Direct Assessments and a multitask model that predicts sentence-level scores and word-level OK/BAD tags. It provides improved correlations with human judgments and increased robustness to critical errors compared to previous metrics.","paper_link":"https://aclanthology.org/2022.wmt-1.52/","implementation_link":null,"verified":false,"created_at":"2026-05-07T16:53:26.585792+00:00","updated_at":"2026-07-05T18:28:03.828508+00:00","statistics":{"total_models":3,"average_score":0.8883333333333333,"min_score":0.887,"max_score":0.89,"score_stddev":0.00152752523165193,"verified_count":0,"self_reported_count":3},"child_benchmarks":[],"linked_dataset":null,"models":[{"rank":1,"model_id":"nova-pro","model_name":"Nova Pro","organization_id":"amazon","organization_name":"Amazon","organization_country":"US","score":0.89,"normalized_score":0.89,"verified":false,"self_reported":true,"self_reported_source":"https://www.amazon.science/publications/the-amazon-nova-family-of-models-technical-report-and-model-card","analysis_method":"COMET22","verification_date":null,"provider_id":null,"input_cost_per_million":null,"output_cost_per_million":null,"context_window":null,"announcement_date":"2024-11-20","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":"nova-lite","model_name":"Nova Lite","organization_id":"amazon","organization_name":"Amazon","organization_country":"US","score":0.888,"normalized_score":0.888,"verified":false,"self_reported":true,"self_reported_source":"https://www.amazon.science/publications/the-amazon-nova-family-of-models-technical-report-and-model-card","analysis_method":"COMET22","verification_date":null,"provider_id":null,"input_cost_per_million":null,"output_cost_per_million":null,"context_window":null,"announcement_date":"2024-11-20","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":3,"model_id":"nova-micro","model_name":"Nova Micro","organization_id":"amazon","organization_name":"Amazon","organization_country":"US","score":0.887,"normalized_score":0.887,"verified":false,"self_reported":true,"self_reported_source":"https://www.amazon.science/publications/the-amazon-nova-family-of-models-technical-report-and-model-card","analysis_method":"COMET22","verification_date":null,"provider_id":null,"input_cost_per_million":null,"output_cost_per_million":null,"context_window":null,"announcement_date":"2024-11-20","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}]}