{"benchmark_id":"apex-agents","benchmark_name":"APEX-Agents","benchmark_description":"APEX-Agents is a benchmark evaluating AI agents on long horizon professional tasks that require sustained reasoning, planning, and execution across complex multi-step workflows.","max_score":1.0,"categories":["reasoning","agents"],"modality":"text","total_models":8,"entries":[{"rank":1,"model_id":"grok-4.6","model_name":"Grok 4.6","organization_name":"xAI","organization_id":"xai","benchmark_score":0.575,"normalized_score":0.575,"verified":false,"self_reported":true,"provider_id":"xai","input_cost_per_million":2.0,"output_cost_per_million":6.0,"speed_rps":null,"context_window":500000,"release_date":"2026-08-12","announcement_date":"2026-08-12","multimodal":true,"param_count":null,"is_new":false},{"rank":2,"model_id":"kimi-k3","model_name":"Kimi K3","organization_name":"Moonshot AI","organization_id":"moonshotai","benchmark_score":0.376,"normalized_score":0.376,"verified":false,"self_reported":true,"provider_id":"novita","input_cost_per_million":3.0,"output_cost_per_million":15.0,"speed_rps":null,"context_window":1048576,"release_date":"2026-07-16","announcement_date":"2026-07-16","multimodal":true,"param_count":2800000000000,"is_new":false},{"rank":3,"model_id":"seed-2.1-pro","model_name":"Seed 2.1 Pro","organization_name":"ByteDance","organization_id":"bytedance","benchmark_score":0.338,"normalized_score":0.338,"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":"2026-06-24","announcement_date":"2026-06-24","multimodal":true,"param_count":null,"is_new":false},{"rank":4,"model_id":"gemini-3.1-pro-preview","model_name":"Gemini 3.1 Pro","organization_name":"Google","organization_id":"google","benchmark_score":0.335,"normalized_score":0.335,"verified":false,"self_reported":true,"provider_id":"google","input_cost_per_million":2.5,"output_cost_per_million":15.0,"speed_rps":90.0,"context_window":1048576,"release_date":"2026-02-19","announcement_date":"2026-02-19","multimodal":true,"param_count":null,"is_new":false},{"rank":5,"model_id":"seed-2.1-turbo","model_name":"Seed 2.1 Turbo","organization_name":"ByteDance","organization_id":"bytedance","benchmark_score":0.292,"normalized_score":0.292,"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":"2026-06-24","announcement_date":"2026-06-24","multimodal":true,"param_count":null,"is_new":false},{"rank":6,"model_id":"kimi-k2.6","model_name":"Kimi K2.6","organization_name":"Moonshot AI","organization_id":"moonshotai","benchmark_score":0.279,"normalized_score":0.279,"verified":false,"self_reported":true,"provider_id":"deepinfra","input_cost_per_million":0.75,"output_cost_per_million":3.5,"speed_rps":null,"context_window":262144,"release_date":"2026-04-20","announcement_date":"2026-04-20","multimodal":true,"param_count":1000000000000,"is_new":false},{"rank":7,"model_id":"minimax-m3","model_name":"MiniMax M3","organization_name":"MiniMax","organization_id":"minimax","benchmark_score":0.277,"normalized_score":0.277,"verified":false,"self_reported":true,"provider_id":"novita","input_cost_per_million":0.3,"output_cost_per_million":1.2,"speed_rps":null,"context_window":1000000,"release_date":"2026-06-01","announcement_date":"2026-06-01","multimodal":true,"param_count":428000000000,"is_new":false},{"rank":8,"model_id":"hy3","model_name":"Hy3","organization_name":"Tencent","organization_id":"tencent","benchmark_score":0.256,"normalized_score":0.256,"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":"2026-07-06","announcement_date":"2026-07-06","multimodal":false,"param_count":295000000000,"is_new":false}]}