{"benchmark_id":"multilingual-mgsm-(cot)","name":"Multilingual MGSM (CoT)","parent_benchmark":null,"categories":["math","reasoning"],"modality":"text","multilingual":true,"max_score":1.0,"language":"en","description":"Multilingual Grade School Math (MGSM) benchmark evaluates language models' chain-of-thought reasoning abilities across ten typologically diverse languages. Contains 250 grade-school math problems manually translated from GSM8K dataset into languages including Bengali and Swahili.","paper_link":"https://arxiv.org/abs/2210.03057","implementation_link":null,"verified":false,"created_at":"2026-05-07T16:53:25.270219+00:00","updated_at":"2026-07-05T18:27:56.467604+00:00","statistics":{"total_models":3,"average_score":0.8246666666666668,"min_score":0.689,"max_score":0.916,"score_stddev":0.11981791741360455,"verified_count":0,"self_reported_count":3},"child_benchmarks":[],"linked_dataset":null,"models":[{"rank":1,"model_id":"llama-3.1-405b-instruct","model_name":"Llama 3.1 405B Instruct","organization_id":"meta","organization_name":"Meta","organization_country":"US","score":0.916,"normalized_score":0.916,"verified":false,"self_reported":true,"self_reported_source":"https://huggingface.co/meta-llama/Meta-Llama-3.1-405B-Instruct","analysis_method":"0-shot, CoT, em","verification_date":null,"provider_id":null,"input_cost_per_million":null,"output_cost_per_million":null,"context_window":null,"announcement_date":"2024-07-23","param_count":405000000000,"is_open_source":true,"is_new":false,"best_latency":null,"latency_provider":null,"best_throughput":null,"throughput_provider":null,"context_provider":null},{"rank":2,"model_id":"llama-3.1-70b-instruct","model_name":"Llama 3.1 70B Instruct","organization_id":"meta","organization_name":"Meta","organization_country":"US","score":0.869,"normalized_score":0.869,"verified":false,"self_reported":true,"self_reported_source":"https://huggingface.co/meta-llama/Meta-Llama-3.1-70B-Instruct","analysis_method":"0-shot Chain-of-Thought","verification_date":null,"provider_id":null,"input_cost_per_million":null,"output_cost_per_million":null,"context_window":null,"announcement_date":"2024-07-23","param_count":70000000000,"is_open_source":true,"is_new":false,"best_latency":null,"latency_provider":null,"best_throughput":null,"throughput_provider":null,"context_provider":null},{"rank":3,"model_id":"llama-3.1-8b-instruct","model_name":"Llama 3.1 8B Instruct","organization_id":"meta","organization_name":"Meta","organization_country":"US","score":0.689,"normalized_score":0.689,"verified":false,"self_reported":true,"self_reported_source":"https://huggingface.co/meta-llama/Meta-Llama-3.1-8B-Instruct","analysis_method":"0-shot","verification_date":null,"provider_id":null,"input_cost_per_million":null,"output_cost_per_million":null,"context_window":null,"announcement_date":"2024-07-23","param_count":8000000000,"is_open_source":true,"is_new":false,"best_latency":null,"latency_provider":null,"best_throughput":null,"throughput_provider":null,"context_provider":null}]}