{"as_of":"2026-10-12","license":"CC-BY-4.0","license_url":"https://creativecommons.org/licenses/by/4.0/","cite_as":"The Local Bench (thelocalbench.com), data as of 2026-10-12, CC BY 4.0","source_url":"https://thelocalbench.com","category":"math","slug":"math-reasoning","label":"Math and reasoning","bench":"AIME 2024, pass@1","unit":"%","higher":true,"src":"Model cards (vendor-reported); sampling settings differ by vendor","groups":[{"label":"AIME 2024, pass@1","ranked":[{"rank":1,"id":"openreasoning-nemotron-32b","name":"OpenReasoning Nemotron 32B","hf_repo":"nvidia/OpenReasoning-Nemotron-32B","score":{"bench":"AIME 2024, pass@1","value":89.2,"unit":"%","flag":"sourced","source":"https://huggingface.co/nvidia/OpenReasoning-Nemotron-32B","note":"AIME24 column, 32B row, model card."}},{"rank":2,"id":"deepseek-r1-0528-qwen3-8b","name":"DeepSeek R1-0528 Qwen3 8B","hf_repo":"deepseek-ai/DeepSeek-R1-0528-Qwen3-8B","score":{"bench":"AIME 2024, pass@1","value":86,"unit":"%","flag":"sourced","source":"https://huggingface.co/deepseek-ai/DeepSeek-R1-0528-Qwen3-8B","note":"AIME 24 column, model card table."}},{"rank":3,"id":"qwq-32b","name":"QwQ 32B","hf_repo":"Qwen/QwQ-32B","score":{"bench":"AIME 2024, pass@1","value":79.5,"unit":"%","flag":"sourced","source":"https://huggingface.co/microsoft/Phi-4-reasoning","note":"Third-party report: the QwQ 32B row in Microsoft's Phi-4-reasoning model card."}},{"rank":4,"id":"phi-4-reasoning","name":"Phi-4 reasoning","hf_repo":"microsoft/Phi-4-reasoning","score":{"bench":"AIME 2024, pass@1","value":75.3,"unit":"%","flag":"sourced","source":"https://huggingface.co/microsoft/Phi-4-reasoning","note":"AIME 24 column, model card table."}},{"rank":5,"id":"openmath-nemotron-7b","name":"OpenMath Nemotron 7B","hf_repo":"nvidia/OpenMath-Nemotron-7B","score":{"bench":"AIME 2024, pass@1","value":74.8,"unit":"%","flag":"sourced","source":"https://huggingface.co/nvidia/OpenMath-Nemotron-7B","note":"AIME24, chain-of-thought pass@1 (the card also gives maj@64 in brackets)."}},{"rank":6,"id":"phi-4-mini-reasoning","name":"Phi-4 mini reasoning","hf_repo":"microsoft/Phi-4-mini-reasoning","score":{"bench":"AIME 2024, pass@1","value":57.5,"unit":"%","flag":"sourced","source":"https://huggingface.co/microsoft/Phi-4-mini-reasoning","note":"\"AIME\" column of the model card (AIME 2024 per the linked paper)."}},{"rank":7,"id":"deepseek-r1-distill-qwen-7b","name":"DeepSeek R1 Distill Qwen 7B","hf_repo":"deepseek-ai/DeepSeek-R1-Distill-Qwen-7B","score":{"bench":"AIME 2024, pass@1","value":55.5,"unit":"%","flag":"sourced","source":"https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-7B","note":"Distilled-model evaluation table, model card."}}],"not_yet_measured":[]}]}