{"benchmark_id":"job-bench","benchmark_name":"Job Bench","benchmark_description":"Job Bench evaluates AI agents on realistic professional tasks that require multi-step planning, research, and production of work artifacts.","max_score":1.0,"categories":["productivity","reasoning","agents"],"modality":"text","total_models":8,"entries":[{"rank":1,"model_id":"muse-spark-1.3","model_name":"Muse Spark 1.3","organization_name":"Meta","organization_id":"meta","benchmark_score":0.649,"normalized_score":0.649,"verified":false,"self_reported":true,"provider_id":"meta","input_cost_per_million":1.25,"output_cost_per_million":4.25,"speed_rps":null,"context_window":1048576,"release_date":"2026-09-02","announcement_date":"2026-09-02","multimodal":true,"param_count":null,"is_new":false},{"rank":2,"model_id":"hy4-preview","model_name":"Hy4 preview","organization_name":"Tencent","organization_id":"tencent","benchmark_score":0.617,"normalized_score":0.617,"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":"2026-08-28","announcement_date":"2026-08-28","multimodal":false,"param_count":770000000000,"is_new":false},{"rank":3,"model_id":"qwen3.8-flash","model_name":"Qwen3.8 Flash","organization_name":"Alibaba Cloud / Qwen Team","organization_id":"qwen","benchmark_score":0.557,"normalized_score":0.557,"verified":false,"self_reported":true,"provider_id":"novita","input_cost_per_million":0.15,"output_cost_per_million":0.47,"speed_rps":null,"context_window":1000000,"release_date":"2026-08-26","announcement_date":"2026-08-26","multimodal":true,"param_count":125000000000,"is_new":false},{"rank":4,"model_id":"qwen3.8-flash-next","model_name":"Qwen3.8-Flash-Next","organization_name":"Alibaba Cloud / Qwen Team","organization_id":"qwen","benchmark_score":0.557,"normalized_score":0.557,"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":"2026-08-26","announcement_date":"2026-08-26","multimodal":true,"param_count":125000000000,"is_new":false},{"rank":5,"model_id":"muse-spark-1.1","model_name":"Muse Spark 1.1","organization_name":"Meta","organization_id":"meta","benchmark_score":0.547,"normalized_score":0.547,"verified":false,"self_reported":true,"provider_id":"meta","input_cost_per_million":1.25,"output_cost_per_million":4.25,"speed_rps":null,"context_window":1048576,"release_date":"2026-07-09","announcement_date":"2026-07-09","multimodal":true,"param_count":null,"is_new":false},{"rank":6,"model_id":"qwen3.8-max","model_name":"Qwen3.8 Max","organization_name":"Alibaba Cloud / Qwen Team","organization_id":"qwen","benchmark_score":0.534,"normalized_score":0.534,"verified":false,"self_reported":true,"provider_id":"deepinfra","input_cost_per_million":1.65,"output_cost_per_million":4.951,"speed_rps":null,"context_window":256000,"release_date":"2026-08-02","announcement_date":"2026-08-02","multimodal":true,"param_count":2400000000000,"is_new":false},{"rank":7,"model_id":"kimi-k3","model_name":"Kimi K3","organization_name":"Moonshot AI","organization_id":"moonshotai","benchmark_score":0.529,"normalized_score":0.529,"verified":false,"self_reported":true,"provider_id":"deepinfra","input_cost_per_million":2.85,"output_cost_per_million":14.25,"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":8,"model_id":"qwen3.8-27b","model_name":"Qwen3.8-27B","organization_name":"Alibaba Cloud / Qwen Team","organization_id":"qwen","benchmark_score":0.334,"normalized_score":0.334,"verified":false,"self_reported":true,"provider_id":"deepinfra","input_cost_per_million":0.4,"output_cost_per_million":3.0,"speed_rps":null,"context_window":262144,"release_date":"2026-08-14","announcement_date":"2026-08-14","multimodal":true,"param_count":27781427952,"is_new":false}]}