{"benchmark_id":"terminal-bench","benchmark_name":"Terminal-Bench","benchmark_description":"Terminal-Bench is a benchmark for testing AI agents in real terminal environments. It evaluates how well agents can handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, security tasks, data science workflows, and cybersecurity vulnerabilities. The benchmark consists of a dataset of ~100 hand-crafted, human-verified tasks and an execution harness that connects language models to a terminal sandbox.","max_score":1.0,"categories":["reasoning","agents","code"],"modality":"text","total_models":25,"entries":[{"rank":1,"model_id":"claude-sonnet-4-5-20250929","model_name":"Claude Sonnet 4.5","organization_name":"Anthropic","organization_id":"anthropic","benchmark_score":0.5,"normalized_score":0.5,"verified":false,"self_reported":true,"provider_id":"anthropic","input_cost_per_million":3.0,"output_cost_per_million":15.0,"speed_rps":42.0,"context_window":200000,"release_date":"2025-09-29","announcement_date":"2025-09-29","multimodal":true,"param_count":null,"is_new":false},{"rank":2,"model_id":"minimax-m2.1","model_name":"MiniMax M2.1","organization_name":"MiniMax","organization_id":"minimax","benchmark_score":0.479,"normalized_score":0.479,"verified":false,"self_reported":true,"provider_id":"minimax","input_cost_per_million":0.3,"output_cost_per_million":1.2,"speed_rps":100.0,"context_window":1000000,"release_date":"2025-12-23","announcement_date":"2025-12-23","multimodal":false,"param_count":230000000000,"is_new":false},{"rank":3,"model_id":"kimi-k2-thinking-0905","model_name":"Kimi K2-Thinking-0905","organization_name":"Moonshot AI","organization_id":"moonshotai","benchmark_score":0.471,"normalized_score":0.471,"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":"2025-09-05","announcement_date":"2025-09-05","multimodal":false,"param_count":1000000000000,"is_new":false},{"rank":4,"model_id":"minimax-m2","model_name":"MiniMax M2","organization_name":"MiniMax","organization_id":"minimax","benchmark_score":0.463,"normalized_score":0.463,"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":204800,"release_date":"2025-10-27","announcement_date":"2025-10-27","multimodal":false,"param_count":230000000000,"is_new":false},{"rank":5,"model_id":"claude-opus-4-1-20250805","model_name":"Claude Opus 4.1","organization_name":"Anthropic","organization_id":"anthropic","benchmark_score":0.433,"normalized_score":0.433,"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":"2025-08-05","announcement_date":"2025-08-05","multimodal":true,"param_count":null,"is_new":false},{"rank":6,"model_id":"nova-2-pro","model_name":"Nova 2 Pro","organization_name":"Amazon","organization_id":"amazon","benchmark_score":0.413,"normalized_score":0.413,"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":"2025-12-02","announcement_date":"2025-12-02","multimodal":true,"param_count":null,"is_new":false},{"rank":7,"model_id":"claude-haiku-4-5-20251001","model_name":"Claude Haiku 4.5","organization_name":"Anthropic","organization_id":"anthropic","benchmark_score":0.41,"normalized_score":0.41,"verified":false,"self_reported":true,"provider_id":"anthropic","input_cost_per_million":1.0,"output_cost_per_million":5.0,"speed_rps":100.0,"context_window":200000,"release_date":"2025-10-15","announcement_date":"2025-10-15","multimodal":true,"param_count":null,"is_new":false},{"rank":8,"model_id":"glm-4.6","model_name":"GLM-4.6","organization_name":"Zhipu AI","organization_id":"zai-org","benchmark_score":0.405,"normalized_score":0.405,"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":"2025-09-30","announcement_date":"2025-09-30","multimodal":true,"param_count":357000000000,"is_new":false},{"rank":9,"model_id":"longcat-flash-chat","model_name":"LongCat-Flash-Chat","organization_name":"Meituan","organization_id":"meituan","benchmark_score":0.3951,"normalized_score":0.3951,"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":"2025-08-29","announcement_date":"2025-08-29","multimodal":false,"param_count":560000000000,"is_new":false},{"rank":10,"model_id":"claude-opus-4-20250514","model_name":"Claude Opus 4","organization_name":"Anthropic","organization_id":"anthropic","benchmark_score":0.392,"normalized_score":0.392,"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":"2025-05-22","announcement_date":"2025-05-22","multimodal":true,"param_count":null,"is_new":false},{"rank":11,"model_id":"deepseek-v3.2-exp","model_name":"DeepSeek-V3.2-Exp","organization_name":"DeepSeek","organization_id":"deepseek","benchmark_score":0.377,"normalized_score":0.377,"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":"2025-09-29","announcement_date":"2025-09-29","multimodal":false,"param_count":685000000000,"is_new":false},{"rank":12,"model_id":"glm-4.5","model_name":"GLM-4.5","organization_name":"Zhipu AI","organization_id":"zai-org","benchmark_score":0.375,"normalized_score":0.375,"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":"2025-07-28","announcement_date":"2025-07-28","multimodal":false,"param_count":355000000000,"is_new":false},{"rank":13,"model_id":"claude-sonnet-4-20250514","model_name":"Claude Sonnet 4","organization_name":"Anthropic","organization_id":"anthropic","benchmark_score":0.355,"normalized_score":0.355,"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":"2025-05-22","announcement_date":"2025-05-22","multimodal":true,"param_count":null,"is_new":false},{"rank":14,"model_id":"claude-3-7-sonnet-20250219","model_name":"Claude 3.7 Sonnet","organization_name":"Anthropic","organization_id":"anthropic","benchmark_score":0.352,"normalized_score":0.352,"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":"2025-02-24","announcement_date":"2025-02-24","multimodal":true,"param_count":null,"is_new":false},{"rank":15,"model_id":"longcat-flash-lite","model_name":"LongCat-Flash-Lite","organization_name":"Meituan","organization_id":"meituan","benchmark_score":0.3375,"normalized_score":0.3375,"verified":false,"self_reported":true,"provider_id":"meituan","input_cost_per_million":0.1,"output_cost_per_million":0.4,"speed_rps":500.0,"context_window":256000,"release_date":"2026-02-05","announcement_date":"2026-02-05","multimodal":false,"param_count":68500000000,"is_new":false},{"rank":16,"model_id":"glm-4.7","model_name":"GLM-4.7","organization_name":"Zhipu AI","organization_id":"zai-org","benchmark_score":0.333,"normalized_score":0.333,"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":"2025-12-22","announcement_date":"2025-12-22","multimodal":true,"param_count":358000000000,"is_new":false},{"rank":17,"model_id":"nova-2-lite","model_name":"Nova 2 Lite","organization_name":"Amazon","organization_id":"amazon","benchmark_score":0.325,"normalized_score":0.325,"verified":false,"self_reported":true,"provider_id":"bedrock","input_cost_per_million":0.3,"output_cost_per_million":2.5,"speed_rps":null,"context_window":1000000,"release_date":"2025-12-02","announcement_date":"2025-12-02","multimodal":true,"param_count":null,"is_new":false},{"rank":18,"model_id":"deepseek-v3.1","model_name":"DeepSeek-V3.1","organization_name":"DeepSeek","organization_id":"deepseek","benchmark_score":0.313,"normalized_score":0.313,"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":"2025-01-10","announcement_date":"2025-01-10","multimodal":false,"param_count":671000000000,"is_new":false},{"rank":19,"model_id":"mimo-v2-flash","model_name":"MiMo-V2-Flash","organization_name":"Xiaomi","organization_id":"xiaomi","benchmark_score":0.305,"normalized_score":0.305,"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":"2025-12-16","announcement_date":"2025-12-16","multimodal":false,"param_count":309000000000,"is_new":false},{"rank":20,"model_id":"glm-4.5-air","model_name":"GLM-4.5-Air","organization_name":"Zhipu AI","organization_id":"zai-org","benchmark_score":0.3,"normalized_score":0.3,"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":"2025-07-28","announcement_date":"2025-07-28","multimodal":false,"param_count":106000000000,"is_new":false},{"rank":21,"model_id":"kimi-k2-instruct","model_name":"Kimi K2 Instruct","organization_name":"Moonshot AI","organization_id":"moonshotai","benchmark_score":0.3,"normalized_score":0.3,"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":"2025-07-11","announcement_date":"2025-07-11","multimodal":false,"param_count":1000000000000,"is_new":false},{"rank":22,"model_id":"nemotron-3-super-120b-a12b","model_name":"Nemotron 3 Super (120B A12B)","organization_name":"NVIDIA","organization_id":"nvidia","benchmark_score":0.2578,"normalized_score":0.2578,"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":"2026-03-11","announcement_date":"2026-03-11","multimodal":false,"param_count":120000000000,"is_new":false},{"rank":23,"model_id":"kimi-k2-instruct-0905","model_name":"Kimi K2-Instruct-0905","organization_name":"Moonshot AI","organization_id":"moonshotai","benchmark_score":0.25,"normalized_score":0.25,"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":"2025-09-05","announcement_date":"2025-09-05","multimodal":false,"param_count":1000000000000,"is_new":false},{"rank":24,"model_id":"nemotron-3-nano-30b-a3b","model_name":"Nemotron 3 Nano (30B A3B)","organization_name":"NVIDIA","organization_id":"nvidia","benchmark_score":0.085,"normalized_score":0.085,"verified":false,"self_reported":true,"provider_id":"deepinfra","input_cost_per_million":0.06,"output_cost_per_million":0.24,"speed_rps":null,"context_window":262144,"release_date":"2025-12-15","announcement_date":"2025-12-15","multimodal":false,"param_count":32000000000,"is_new":false},{"rank":25,"model_id":"deepseek-r1-0528","model_name":"DeepSeek-R1-0528","organization_name":"DeepSeek","organization_id":"deepseek","benchmark_score":0.057,"normalized_score":0.057,"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":"2025-05-28","announcement_date":"2025-05-28","multimodal":false,"param_count":671000000000,"is_new":false}]}