{"benchmark_id":"browsecomp-vl","benchmark_name":"BrowseComp-VL","benchmark_description":"BrowseComp-VL is the vision-language variant of BrowseComp, evaluating multimodal models on web browsing comprehension tasks that require processing visual web page content alongside text.","max_score":1.0,"categories":["multimodal","search","agents","vision"],"modality":"multimodal","total_models":1,"entries":[{"rank":1,"model_id":"glm-5v-turbo","model_name":"GLM-5V-Turbo","organization_name":"Zhipu AI","organization_id":"zai-org","benchmark_score":0.519,"normalized_score":0.519,"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":"2026-04-02","announcement_date":"2026-04-02","multimodal":true,"param_count":null,"is_new":false}]}