{"benchmark_id":"drop","benchmark_name":"DROP","benchmark_description":"DROP (Discrete Reasoning Over Paragraphs) is a reading comprehension benchmark requiring discrete reasoning over paragraph content. It contains crowdsourced, adversarially-created questions that require resolving references and performing discrete operations like addition, counting, or sorting, demanding comprehensive paragraph understanding beyond paraphrase-and-entity-typing shortcuts.","max_score":1.0,"categories":["math","reasoning"],"modality":"text","total_models":30,"entries":[{"rank":1,"model_id":"deepseek-v3","model_name":"DeepSeek-V3","organization_name":"DeepSeek","organization_id":"deepseek","benchmark_score":0.916,"normalized_score":0.916,"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":"2024-12-25","announcement_date":"2024-12-25","multimodal":false,"param_count":671000000000,"is_new":false},{"rank":2,"model_id":"claude-3-5-sonnet-20241022","model_name":"Claude 3.5 Sonnet","organization_name":"Anthropic","organization_id":"anthropic","benchmark_score":0.871,"normalized_score":0.871,"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":"2024-10-22","announcement_date":"2024-10-22","multimodal":true,"param_count":null,"is_new":false},{"rank":3,"model_id":"claude-3-5-sonnet-20240620","model_name":"Claude 3.5 Sonnet","organization_name":"Anthropic","organization_id":"anthropic","benchmark_score":0.871,"normalized_score":0.871,"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":"2024-06-21","announcement_date":"2024-06-21","multimodal":true,"param_count":null,"is_new":false},{"rank":4,"model_id":"mimo-v2.5-pro","model_name":"MiMo-V2.5-Pro","organization_name":"Xiaomi","organization_id":"xiaomi","benchmark_score":0.863,"normalized_score":0.863,"verified":false,"self_reported":true,"provider_id":"xiaomi","input_cost_per_million":0.435,"output_cost_per_million":0.87,"speed_rps":null,"context_window":1048576,"release_date":"2026-04-27","announcement_date":"2026-04-27","multimodal":false,"param_count":1023244718976,"is_new":false},{"rank":5,"model_id":"gpt-4-turbo-2024-04-09","model_name":"GPT-4 Turbo","organization_name":"OpenAI","organization_id":"openai","benchmark_score":0.86,"normalized_score":0.86,"verified":false,"self_reported":true,"provider_id":"azure","input_cost_per_million":10.0,"output_cost_per_million":30.0,"speed_rps":97.0,"context_window":128000,"release_date":"2024-04-09","announcement_date":"2024-04-09","multimodal":false,"param_count":null,"is_new":false},{"rank":6,"model_id":"nova-pro","model_name":"Nova Pro","organization_name":"Amazon","organization_id":"amazon","benchmark_score":0.854,"normalized_score":0.854,"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":"2024-11-20","announcement_date":"2024-11-20","multimodal":true,"param_count":null,"is_new":false},{"rank":7,"model_id":"llama-3.1-405b-instruct","model_name":"Llama 3.1 405B Instruct","organization_name":"Meta","organization_id":"meta","benchmark_score":0.848,"normalized_score":0.848,"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":"2024-07-23","announcement_date":"2024-07-23","multimodal":false,"param_count":405000000000,"is_new":false},{"rank":8,"model_id":"gpt-4o-2024-05-13","model_name":"GPT-4o","organization_name":"OpenAI","organization_id":"openai","benchmark_score":0.834,"normalized_score":0.834,"verified":false,"self_reported":true,"provider_id":"openai","input_cost_per_million":2.5,"output_cost_per_million":10.0,"speed_rps":100.0,"context_window":128000,"release_date":"2024-05-13","announcement_date":"2024-05-13","multimodal":true,"param_count":null,"is_new":false},{"rank":9,"model_id":"claude-3-5-haiku-20241022","model_name":"Claude 3.5 Haiku","organization_name":"Anthropic","organization_id":"anthropic","benchmark_score":0.831,"normalized_score":0.831,"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":"2024-10-22","announcement_date":"2024-10-22","multimodal":false,"param_count":null,"is_new":false},{"rank":10,"model_id":"claude-3-opus-20240229","model_name":"Claude 3 Opus","organization_name":"Anthropic","organization_id":"anthropic","benchmark_score":0.831,"normalized_score":0.831,"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":"2024-02-29","announcement_date":"2024-02-29","multimodal":true,"param_count":null,"is_new":false},{"rank":11,"model_id":"gpt-4-0613","model_name":"GPT-4","organization_name":"OpenAI","organization_id":"openai","benchmark_score":0.809,"normalized_score":0.809,"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":"2023-06-13","announcement_date":"2023-06-13","multimodal":true,"param_count":null,"is_new":false},{"rank":12,"model_id":"nova-lite","model_name":"Nova Lite","organization_name":"Amazon","organization_id":"amazon","benchmark_score":0.802,"normalized_score":0.802,"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":"2024-11-20","announcement_date":"2024-11-20","multimodal":true,"param_count":null,"is_new":false},{"rank":13,"model_id":"gpt-4o-mini-2024-07-18","model_name":"GPT-4o mini","organization_name":"OpenAI","organization_id":"openai","benchmark_score":0.797,"normalized_score":0.797,"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":"2024-07-18","announcement_date":"2024-07-18","multimodal":true,"param_count":null,"is_new":false},{"rank":14,"model_id":"llama-3.1-70b-instruct","model_name":"Llama 3.1 70B Instruct","organization_name":"Meta","organization_id":"meta","benchmark_score":0.796,"normalized_score":0.796,"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":"2024-07-23","announcement_date":"2024-07-23","multimodal":false,"param_count":70000000000,"is_new":false},{"rank":15,"model_id":"nova-micro","model_name":"Nova Micro","organization_name":"Amazon","organization_id":"amazon","benchmark_score":0.793,"normalized_score":0.793,"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":"2024-11-20","announcement_date":"2024-11-20","multimodal":false,"param_count":null,"is_new":false},{"rank":16,"model_id":"longcat-flash-chat","model_name":"LongCat-Flash-Chat","organization_name":"Meituan","organization_id":"meituan","benchmark_score":0.7906,"normalized_score":0.7906,"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":"2025-08-29","announcement_date":"2025-08-29","multimodal":false,"param_count":560000000000,"is_new":false},{"rank":17,"model_id":"claude-3-sonnet-20240229","model_name":"Claude 3 Sonnet","organization_name":"Anthropic","organization_id":"anthropic","benchmark_score":0.789,"normalized_score":0.789,"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":"2024-02-29","announcement_date":"2024-02-29","multimodal":true,"param_count":null,"is_new":false},{"rank":18,"model_id":"claude-3-haiku-20240307","model_name":"Claude 3 Haiku","organization_name":"Anthropic","organization_id":"anthropic","benchmark_score":0.784,"normalized_score":0.784,"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":"2024-03-13","announcement_date":"2024-03-13","multimodal":true,"param_count":null,"is_new":false},{"rank":19,"model_id":"phi-4","model_name":"Phi 4","organization_name":"Microsoft","organization_id":"microsoft","benchmark_score":0.755,"normalized_score":0.755,"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":"2024-12-12","announcement_date":"2024-12-12","multimodal":false,"param_count":14700000000,"is_new":false},{"rank":20,"model_id":"gemini-1.5-pro","model_name":"Gemini 1.5 Pro","organization_name":"Google","organization_id":"google","benchmark_score":0.749,"normalized_score":0.749,"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":"2024-05-01","announcement_date":"2024-05-01","multimodal":true,"param_count":null,"is_new":false},{"rank":21,"model_id":"gpt-3.5-turbo-0125","model_name":"GPT-3.5 Turbo","organization_name":"OpenAI","organization_id":"openai","benchmark_score":0.702,"normalized_score":0.702,"verified":false,"self_reported":false,"provider_id":"openai","input_cost_per_million":0.5,"output_cost_per_million":1.5,"speed_rps":100.0,"context_window":16385,"release_date":"2023-03-21","announcement_date":"2023-03-21","multimodal":false,"param_count":null,"is_new":false},{"rank":22,"model_id":"gemma-3n-e4b","model_name":"Gemma 3n E4B","organization_name":"Google","organization_id":"google","benchmark_score":0.608,"normalized_score":0.608,"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":"2025-06-26","announcement_date":"2025-06-26","multimodal":true,"param_count":8000000000,"is_new":false},{"rank":23,"model_id":"gemma-3n-e4b-it-litert-preview","model_name":"Gemma 3n E4B Instructed LiteRT Preview","organization_name":"Google","organization_id":"google","benchmark_score":0.608,"normalized_score":0.608,"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":"2025-05-20","announcement_date":"2025-05-20","multimodal":true,"param_count":1910000000,"is_new":false},{"rank":24,"model_id":"llama-3.1-8b-instruct","model_name":"Llama 3.1 8B Instruct","organization_name":"Meta","organization_id":"meta","benchmark_score":0.595,"normalized_score":0.595,"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":"2024-07-23","announcement_date":"2024-07-23","multimodal":false,"param_count":8000000000,"is_new":false},{"rank":25,"model_id":"granite-3.3-8b-instruct","model_name":"Granite 3.3 8B Instruct","organization_name":"IBM","organization_id":"ibm","benchmark_score":0.5936,"normalized_score":0.5936,"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":"2025-04-16","announcement_date":"2025-04-16","multimodal":true,"param_count":8000000000,"is_new":false},{"rank":26,"model_id":"gemma-3n-e2b","model_name":"Gemma 3n E2B","organization_name":"Google","organization_id":"google","benchmark_score":0.539,"normalized_score":0.539,"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":"2025-06-26","announcement_date":"2025-06-26","multimodal":true,"param_count":8000000000,"is_new":false},{"rank":27,"model_id":"gemma-3n-e2b-it-litert-preview","model_name":"Gemma 3n E2B Instructed LiteRT (Preview)","organization_name":"Google","organization_id":"google","benchmark_score":0.539,"normalized_score":0.539,"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":"2025-05-20","announcement_date":"2025-05-20","multimodal":true,"param_count":1910000000,"is_new":false},{"rank":28,"model_id":"granite-4.0-tiny-preview","model_name":"IBM Granite 4.0 Tiny Preview","organization_name":"IBM","organization_id":"ibm","benchmark_score":0.462,"normalized_score":0.462,"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":"2025-05-02","announcement_date":"2025-05-02","multimodal":false,"param_count":7000000000,"is_new":false},{"rank":29,"model_id":"granite-3.3-8b-base","model_name":"Granite 3.3 8B Base","organization_name":"IBM","organization_id":"ibm","benchmark_score":0.3614,"normalized_score":0.3614,"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":"2025-04-16","announcement_date":"2025-04-16","multimodal":true,"param_count":8170000000,"is_new":false},{"rank":30,"model_id":"ernie-4.5","model_name":"ERNIE 4.5","organization_name":"Baidu","organization_id":"baidu","benchmark_score":0.286,"normalized_score":0.286,"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":"2025-06-25","announcement_date":"2025-06-25","multimodal":false,"param_count":21000000000,"is_new":false}]}