{"benchmark_id":"healthbench-professional","benchmark_name":"HealthBench Professional","benchmark_description":"HealthBench Professional evaluates model capability and safety for clinician use cases using real clinician-style chats and physician-authored grading rubrics.","max_score":1.0,"categories":["healthcare"],"modality":"text","total_models":17,"entries":[{"rank":1,"model_id":"claude-sonnet-5-5","model_name":"Claude Sonnet 5.5","organization_name":"Anthropic","organization_id":"anthropic","benchmark_score":0.692,"normalized_score":0.692,"verified":false,"self_reported":true,"provider_id":"anthropic","input_cost_per_million":2.0,"output_cost_per_million":10.0,"speed_rps":42.0,"context_window":1000000,"release_date":"2026-09-28","announcement_date":"2026-09-28","multimodal":true,"param_count":null,"is_new":false},{"rank":2,"model_id":"claude-fable-5","model_name":"Claude Fable 5","organization_name":"Anthropic","organization_id":"anthropic","benchmark_score":0.66,"normalized_score":0.66,"verified":false,"self_reported":true,"provider_id":"anthropic","input_cost_per_million":10.0,"output_cost_per_million":50.0,"speed_rps":null,"context_window":1000000,"release_date":"2026-06-09","announcement_date":"2026-06-09","multimodal":true,"param_count":null,"is_new":false},{"rank":3,"model_id":"claude-opus-5-5","model_name":"Claude Opus 5.5","organization_name":"Anthropic","organization_id":"anthropic","benchmark_score":0.656,"normalized_score":0.656,"verified":false,"self_reported":true,"provider_id":"anthropic","input_cost_per_million":4.0,"output_cost_per_million":20.0,"speed_rps":null,"context_window":1000000,"release_date":"2026-09-22","announcement_date":"2026-09-22","multimodal":true,"param_count":null,"is_new":false},{"rank":4,"model_id":"claude-haiku-5-5","model_name":"Claude Haiku 5.5","organization_name":"Anthropic","organization_id":"anthropic","benchmark_score":0.648,"normalized_score":0.648,"verified":false,"self_reported":true,"provider_id":"anthropic","input_cost_per_million":0.1,"output_cost_per_million":0.5,"speed_rps":null,"context_window":1000000,"release_date":"2026-10-07","announcement_date":"2026-10-07","multimodal":true,"param_count":null,"is_new":true},{"rank":5,"model_id":"gpt-6.1-sol","model_name":"GPT-6.1 Sol","organization_name":"OpenAI","organization_id":"openai","benchmark_score":0.642,"normalized_score":0.642,"verified":false,"self_reported":true,"provider_id":"openai","input_cost_per_million":2.0,"output_cost_per_million":10.0,"speed_rps":null,"context_window":1050000,"release_date":"2026-09-29","announcement_date":"2026-09-29","multimodal":true,"param_count":null,"is_new":false},{"rank":6,"model_id":"gpt-6-astra","model_name":"GPT-6 Astra","organization_name":"OpenAI","organization_id":"openai","benchmark_score":0.634,"normalized_score":0.634,"verified":false,"self_reported":true,"provider_id":"openai","input_cost_per_million":10.0,"output_cost_per_million":50.0,"speed_rps":null,"context_window":1050000,"release_date":"2026-09-04","announcement_date":"2026-09-03","multimodal":true,"param_count":null,"is_new":false},{"rank":7,"model_id":"muse-wonder-full","model_name":"Muse Wonder Full","organization_name":"Meta","organization_id":"meta","benchmark_score":0.623,"normalized_score":0.623,"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":"2026-10-12","announcement_date":"2026-10-12","multimodal":true,"param_count":null,"is_new":true},{"rank":8,"model_id":"muse-wonder","model_name":"Muse Wonder","organization_name":"Meta","organization_id":"meta","benchmark_score":0.605,"normalized_score":0.605,"verified":false,"self_reported":true,"provider_id":"meta","input_cost_per_million":4.0,"output_cost_per_million":20.0,"speed_rps":null,"context_window":1048576,"release_date":"2026-10-12","announcement_date":"2026-10-12","multimodal":true,"param_count":null,"is_new":true},{"rank":9,"model_id":"gpt-5.6-sol","model_name":"GPT-5.6 Sol","organization_name":"OpenAI","organization_id":"openai","benchmark_score":0.605,"normalized_score":0.605,"verified":false,"self_reported":true,"provider_id":"openai","input_cost_per_million":5.0,"output_cost_per_million":30.0,"speed_rps":null,"context_window":1050000,"release_date":"2026-07-09","announcement_date":"2026-07-09","multimodal":true,"param_count":null,"is_new":false},{"rank":10,"model_id":"claude-opus-5","model_name":"Claude Opus 5","organization_name":"Anthropic","organization_id":"anthropic","benchmark_score":0.598,"normalized_score":0.598,"verified":false,"self_reported":true,"provider_id":"anthropic","input_cost_per_million":5.0,"output_cost_per_million":25.0,"speed_rps":null,"context_window":1000000,"release_date":"2026-07-24","announcement_date":"2026-07-24","multimodal":true,"param_count":null,"is_new":false},{"rank":11,"model_id":"claude-sonnet-5","model_name":"Claude Sonnet 5","organization_name":"Anthropic","organization_id":"anthropic","benchmark_score":0.578,"normalized_score":0.578,"verified":false,"self_reported":true,"provider_id":"anthropic","input_cost_per_million":2.0,"output_cost_per_million":10.0,"speed_rps":42.0,"context_window":1000000,"release_date":"2026-06-30","announcement_date":"2026-06-30","multimodal":true,"param_count":null,"is_new":false},{"rank":12,"model_id":"gpt-5.6-terra","model_name":"GPT-5.6 Terra","organization_name":"OpenAI","organization_id":"openai","benchmark_score":0.577,"normalized_score":0.577,"verified":false,"self_reported":true,"provider_id":"openai","input_cost_per_million":2.0,"output_cost_per_million":12.0,"speed_rps":null,"context_window":1050000,"release_date":"2026-07-09","announcement_date":"2026-07-09","multimodal":true,"param_count":null,"is_new":false},{"rank":13,"model_id":"grok-4.7","model_name":"Grok 4.7","organization_name":"xAI","organization_id":"xai","benchmark_score":0.567,"normalized_score":0.567,"verified":false,"self_reported":true,"provider_id":"xai","input_cost_per_million":2.0,"output_cost_per_million":6.0,"speed_rps":null,"context_window":500000,"release_date":"2026-09-21","announcement_date":"2026-09-21","multimodal":true,"param_count":null,"is_new":false},{"rank":14,"model_id":"claude-opus-4-8","model_name":"Claude Opus 4.8","organization_name":"Anthropic","organization_id":"anthropic","benchmark_score":0.558,"normalized_score":0.558,"verified":false,"self_reported":true,"provider_id":"deepinfra","input_cost_per_million":5.0,"output_cost_per_million":25.0,"speed_rps":null,"context_window":1000000,"release_date":"2026-05-28","announcement_date":"2026-05-28","multimodal":true,"param_count":null,"is_new":false},{"rank":15,"model_id":"gpt-5.6-luna","model_name":"GPT-5.6 Luna","organization_name":"OpenAI","organization_id":"openai","benchmark_score":0.557,"normalized_score":0.557,"verified":false,"self_reported":true,"provider_id":"openai","input_cost_per_million":0.2,"output_cost_per_million":1.2,"speed_rps":null,"context_window":1050000,"release_date":"2026-07-09","announcement_date":"2026-07-09","multimodal":true,"param_count":null,"is_new":false},{"rank":16,"model_id":"gpt-5.5-instant","model_name":"GPT-5.5 Instant","organization_name":"OpenAI","organization_id":"openai","benchmark_score":0.384,"normalized_score":0.384,"verified":false,"self_reported":true,"provider_id":"openai","input_cost_per_million":5.0,"output_cost_per_million":30.0,"speed_rps":100.0,"context_window":400000,"release_date":"2026-05-05","announcement_date":"2026-05-05","multimodal":true,"param_count":null,"is_new":false},{"rank":17,"model_id":"mai-thinking-1","model_name":"MAI-Thinking-1","organization_name":"Microsoft","organization_id":"microsoft","benchmark_score":0.35,"normalized_score":0.35,"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":"2026-06-02","announcement_date":"2026-06-02","multimodal":false,"param_count":1000000000000,"is_new":false}]}