{"benchmark_id":"mmau","benchmark_name":"MMAU","benchmark_description":"A massive multi-task audio understanding and reasoning benchmark comprising 10,000 carefully curated audio clips paired with human-annotated natural language questions spanning speech, environmental sounds, and music. Requires expert-level knowledge and complex reasoning across 27 distinct skills.","max_score":1.0,"categories":["multimodal","reasoning","audio"],"modality":"multimodal","total_models":3,"entries":[{"rank":1,"model_id":"inkling-small","model_name":"Inkling-Small","organization_name":"Thinking Machines Lab","organization_id":"thinking-machines","benchmark_score":0.77,"normalized_score":0.77,"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":2,"model_id":"nova-2-omni","model_name":"Nova 2 Omni","organization_name":"Amazon","organization_id":"amazon","benchmark_score":0.753,"normalized_score":0.753,"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":"2025-12-02","announcement_date":"2025-12-02","multimodal":true,"param_count":null,"is_new":false},{"rank":3,"model_id":"qwen2.5-omni-7b","model_name":"Qwen2.5-Omni-7B","organization_name":"Alibaba Cloud / Qwen Team","organization_id":"qwen","benchmark_score":0.656,"normalized_score":0.656,"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":"2025-03-27","announcement_date":"2025-03-27","multimodal":true,"param_count":7000000000,"is_new":false}]}