{"benchmark_id":"aime-2026","benchmark_name":"AIME 2026","benchmark_description":"All 30 problems from the 2026 American Invitational Mathematics Examination (AIME I and AIME II), testing olympiad-level mathematical reasoning with integer answers from 000-999. Used as an AI benchmark to evaluate large language models' ability to solve complex mathematical problems requiring multi-step logical deductions and structured symbolic reasoning.","max_score":1.0,"categories":["math","reasoning"],"modality":"text","total_models":22,"entries":[{"rank":1,"model_id":"glm-5.2","model_name":"GLM-5.2","organization_name":"Zhipu AI","organization_id":"zai-org","benchmark_score":0.992,"normalized_score":0.992,"verified":false,"self_reported":true,"provider_id":"deepinfra","input_cost_per_million":0.95,"output_cost_per_million":3.0,"speed_rps":null,"context_window":1048576,"release_date":"2026-06-16","announcement_date":"2026-06-16","multimodal":false,"param_count":753000000000,"is_new":false},{"rank":2,"model_id":"sakana-namazu","model_name":"Sakana Namazu","organization_name":"Sakana AI","organization_id":"sakana","benchmark_score":0.9667,"normalized_score":0.9667,"verified":false,"self_reported":true,"provider_id":"sakana","input_cost_per_million":0.95,"output_cost_per_million":4.0,"speed_rps":null,"context_window":256000,"release_date":"2026-08-03","announcement_date":"2026-08-03","multimodal":true,"param_count":null,"is_new":false},{"rank":3,"model_id":"kimi-k2.6","model_name":"Kimi K2.6","organization_name":"Moonshot AI","organization_id":"moonshotai","benchmark_score":0.964,"normalized_score":0.964,"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":4,"model_id":"inkling-small","model_name":"Inkling-Small","organization_name":"Thinking Machines Lab","organization_id":"thinking-machines","benchmark_score":0.955,"normalized_score":0.955,"verified":false,"self_reported":true,"provider_id":"thinking-machines","input_cost_per_million":0.3,"output_cost_per_million":1.2,"speed_rps":null,"context_window":256000,"release_date":"2026-07-30","announcement_date":"2026-07-30","multimodal":true,"param_count":276000000000,"is_new":false},{"rank":5,"model_id":"solar-pro4","model_name":"Solar Pro 4","organization_name":"Upstage","organization_id":"upstage","benchmark_score":0.953,"normalized_score":0.953,"verified":false,"self_reported":true,"provider_id":"upstage","input_cost_per_million":0.3,"output_cost_per_million":1.2,"speed_rps":null,"context_window":524288,"release_date":"2026-08-06","announcement_date":"2026-08-06","multimodal":false,"param_count":null,"is_new":false},{"rank":6,"model_id":"glm-5.1","model_name":"GLM-5.1","organization_name":"Zhipu AI","organization_id":"zai-org","benchmark_score":0.953,"normalized_score":0.953,"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":7,"model_id":"qwen3.6-plus","model_name":"Qwen3.6 Plus","organization_name":"Alibaba Cloud / Qwen Team","organization_id":"qwen","benchmark_score":0.953,"normalized_score":0.953,"verified":false,"self_reported":true,"provider_id":"together","input_cost_per_million":0.5,"output_cost_per_million":3.0,"speed_rps":null,"context_window":1000000,"release_date":"2026-03-31","announcement_date":"2026-04-02","multimodal":true,"param_count":null,"is_new":false},{"rank":8,"model_id":"muse-glimmer-30b","model_name":"Muse Glimmer-30B","organization_name":"Meta","organization_id":"meta","benchmark_score":0.947,"normalized_score":0.947,"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":"2026-08-10","announcement_date":"2026-08-10","multimodal":true,"param_count":29600000000,"is_new":false},{"rank":9,"model_id":"mai-thinking-1","model_name":"MAI-Thinking-1","organization_name":"Microsoft","organization_id":"microsoft","benchmark_score":0.945,"normalized_score":0.945,"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":"2026-06-02","announcement_date":"2026-06-02","multimodal":false,"param_count":1000000000000,"is_new":false},{"rank":10,"model_id":"seed-2.0-pro","model_name":"Seed 2.0 Pro","organization_name":"ByteDance","organization_id":"bytedance","benchmark_score":0.942,"normalized_score":0.942,"verified":false,"self_reported":true,"provider_id":"deepinfra","input_cost_per_million":0.5,"output_cost_per_million":3.0,"speed_rps":null,"context_window":256000,"release_date":"2026-02-14","announcement_date":"2026-02-14","multimodal":true,"param_count":null,"is_new":false},{"rank":11,"model_id":"qwen3.6-27b","model_name":"Qwen3.6-27B","organization_name":"Alibaba Cloud / Qwen Team","organization_id":"qwen","benchmark_score":0.941,"normalized_score":0.941,"verified":false,"self_reported":true,"provider_id":"novita","input_cost_per_million":0.6,"output_cost_per_million":3.6,"speed_rps":null,"context_window":262144,"release_date":"2026-04-21","announcement_date":"2026-04-21","multimodal":true,"param_count":27781427952,"is_new":false},{"rank":12,"model_id":"qwen3.6-35b-a3b","model_name":"Qwen3.6-35B-A3B","organization_name":"Alibaba Cloud / Qwen Team","organization_id":"qwen","benchmark_score":0.927,"normalized_score":0.927,"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":"2026-04-16","announcement_date":"2026-04-16","multimodal":true,"param_count":35000000000,"is_new":false},{"rank":13,"model_id":"exaone-4.5-33b","model_name":"EXAONE 4.5 33B","organization_name":"LG AI Research","organization_id":"lg","benchmark_score":0.926,"normalized_score":0.926,"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":"2026-04-09","announcement_date":"2026-04-09","multimodal":true,"param_count":33000000000,"is_new":false},{"rank":14,"model_id":"mai-code-1-flash","model_name":"MAI-Code-1-Flash","organization_name":"Microsoft","organization_id":"microsoft","benchmark_score":0.925,"normalized_score":0.925,"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":"2026-06-02","announcement_date":"2026-06-02","multimodal":false,"param_count":null,"is_new":false},{"rank":15,"model_id":"qwen3.5-397b-a17b","model_name":"Qwen3.5-397B-A17B","organization_name":"Alibaba Cloud / Qwen Team","organization_id":"qwen","benchmark_score":0.913,"normalized_score":0.913,"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":"2026-02-16","announcement_date":"2026-02-16","multimodal":true,"param_count":397000000000,"is_new":false},{"rank":16,"model_id":"gemma-4-31b-it","model_name":"Gemma 4 31B","organization_name":"Google","organization_id":"google","benchmark_score":0.892,"normalized_score":0.892,"verified":false,"self_reported":true,"provider_id":"deepinfra","input_cost_per_million":0.13,"output_cost_per_million":0.38,"speed_rps":null,"context_window":262144,"release_date":"2026-04-02","announcement_date":"2026-04-02","multimodal":true,"param_count":30700000000,"is_new":false},{"rank":17,"model_id":"gemma-4-26b-a4b-it","model_name":"Gemma 4 26B-A4B","organization_name":"Google","organization_id":"google","benchmark_score":0.883,"normalized_score":0.883,"verified":false,"self_reported":true,"provider_id":"novita","input_cost_per_million":0.13,"output_cost_per_million":0.4,"speed_rps":null,"context_window":262144,"release_date":"2026-04-02","announcement_date":"2026-04-02","multimodal":true,"param_count":25200000000,"is_new":false},{"rank":18,"model_id":"seed-2.0-lite","model_name":"Seed 2.0 Lite","organization_name":"ByteDance","organization_id":"bytedance","benchmark_score":0.883,"normalized_score":0.883,"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":"2026-02-14","announcement_date":"2026-02-14","multimodal":true,"param_count":null,"is_new":false},{"rank":19,"model_id":"gemma-4-12b-it","model_name":"Gemma 4 12B","organization_name":"Google","organization_id":"google","benchmark_score":0.775,"normalized_score":0.775,"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":"2026-05-23","announcement_date":"2026-05-23","multimodal":true,"param_count":11959730224,"is_new":false},{"rank":20,"model_id":"diffusiongemma-26b-a4b-it","model_name":"DiffusionGemma 26B-A4B","organization_name":"Google","organization_id":"google","benchmark_score":0.691,"normalized_score":0.691,"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":"2026-06-10","announcement_date":"2026-06-10","multimodal":true,"param_count":25200000000,"is_new":false},{"rank":21,"model_id":"gemma-4-e4b-it","model_name":"Gemma 4 E4B","organization_name":"Google","organization_id":"google","benchmark_score":0.425,"normalized_score":0.425,"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":8000000000,"is_new":false},{"rank":22,"model_id":"gemma-4-e2b-it","model_name":"Gemma 4 E2B","organization_name":"Google","organization_id":"google","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":"2026-04-02","announcement_date":"2026-04-02","multimodal":true,"param_count":5100000000,"is_new":false}]}