{"benchmark_id":"vending-bench-2","benchmark_name":"Vending-Bench 2","benchmark_description":"Vending-Bench 2 tests longer horizon planning capabilities by evaluating how well AI models can manage a simulated vending machine business over extended periods. The benchmark measures a model's ability to maintain consistent tool usage and decision-making for a full simulated year of operation, driving higher returns without drifting off task.","max_score":1.0,"categories":["reasoning","agents"],"modality":"text","total_models":4,"entries":[{"rank":1,"model_id":"claude-opus-4-6","model_name":"Claude Opus 4.6","organization_name":"Anthropic","organization_id":"anthropic","benchmark_score":8017.59,"normalized_score":null,"verified":false,"self_reported":true,"provider_id":"anthropic","input_cost_per_million":5.0,"output_cost_per_million":25.0,"speed_rps":42.0,"context_window":1000000,"release_date":"2026-02-05","announcement_date":"2026-02-05","multimodal":true,"param_count":null,"is_new":false},{"rank":2,"model_id":"glm-5.1","model_name":"GLM-5.1","organization_name":"Zhipu AI","organization_id":"zai-org","benchmark_score":5634.41,"normalized_score":null,"verified":false,"self_reported":true,"provider_id":"friendli","input_cost_per_million":1.4,"output_cost_per_million":4.4,"speed_rps":null,"context_window":200000,"release_date":"2026-04-07","announcement_date":"2026-04-07","multimodal":false,"param_count":754000000000,"is_new":false},{"rank":3,"model_id":"gemini-3-pro-preview","model_name":"Gemini 3 Pro","organization_name":"Google","organization_id":"google","benchmark_score":5478.16,"normalized_score":null,"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":"2025-11-18","announcement_date":"2025-11-18","multimodal":true,"param_count":null,"is_new":false},{"rank":4,"model_id":"gemini-3-flash-preview","model_name":"Gemini 3 Flash","organization_name":"Google","organization_id":"google","benchmark_score":3635.0,"normalized_score":null,"verified":false,"self_reported":true,"provider_id":"google","input_cost_per_million":0.5,"output_cost_per_million":3.0,"speed_rps":null,"context_window":1000000,"release_date":"2025-12-17","announcement_date":"2025-12-17","multimodal":true,"param_count":null,"is_new":false}]}