{"benchmark_id":"swe-mm","benchmark_name":"SWE-MM","benchmark_description":"SWE-MM evaluates software-engineering agents on repository tasks that require understanding both source code and visual evidence.","max_score":1.0,"categories":["multimodal","agents","code","vision"],"modality":"multimodal","total_models":1,"entries":[{"rank":1,"model_id":"qwen3.8-27b","model_name":"Qwen3.8-27B","organization_name":"Alibaba Cloud / Qwen Team","organization_id":"qwen","benchmark_score":0.386,"normalized_score":0.386,"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":"2026-08-14","announcement_date":"2026-08-14","multimodal":true,"param_count":27781427952,"is_new":false}]}