MiMo-Embodied: X-Embodied Foundation Model

MiMo-Embodied: X-Embodied Foundation Model | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Biological Sciences - Article MiMo-Embodied: X-Embodied Foundation Model Xiaoshuai Hao, Lei Zhou, Zhijian Huang, Zhiwen Hou, Yingbo Tang, and 39 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8493355/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract MiMo-Embodied is a groundbreaking cross-embodied foundation model that integrates indoor robotics and outdoor autonomous driving in one model, effectively addressing the historical ''domain gap'' that has siloed Embodied AI and driving systems. The model achieves state-of-the-art performance across 29 benchmarks, comprising 17 tasks in Embodied AI (including task planning, affordance prediction and spatial understanding) and 12 in autonomous driving (spanning environmental perception, status prediction, and driving planning). Central to these gains is a multi-stage training strategy that couples curated general and domain-specific data with supervised alignment, chain-of-thought reasoning supervision, and reinforcement learning, yielding robust positive transfer and mutual reinforcement between embodiments. The model consistently outperforms specialized, open-source, and closed-source counterparts while retaining general visual semantic understanding. These results indicate that a single vision-language foundation model can acquire cohesive physical intelligence across diverse embodiments and environments, suggesting that carefully staged, multimodal training is sufficient to unlock cross-task generalization without architectural specialization. This work offers a scalable framework for developing unified embodied systems. Code and models are available at https://github.com/XiaomiMiMo/MiMo-Embodied . Biological sciences/Biological techniques/Software Scientific community and society/Scientific community Cross-Embodied Foundation Models Embodied AI Autonomous Driving Full Text Additional Declarations There is NO Competing Interest. Cite Share Download PDF Status: Under Review Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8493355","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Biological Sciences - Article","associatedPublications":[],"authors":[{"id":569015866,"identity":"1f9cb9f2-9f20-4bc6-9edb-2375ae28dac9","order_by":0,"name":"Xiaoshuai Hao","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA+0lEQVRIiWNgGAWjYBACAxCRAGUc/mPAJsfG3n6AeC0HeAr4jPl4ziQQ1gJn8HyQS5wn4WCAWz0QmLMfPibxoMbG3lwi+eEBCQOz9DYJoK0/Krbh1GLZk5ZskHAsLXHnjDSDAwYGablt0o0HGHvO3MbtsAM5hg8S2A4nGNxIMDiQYHAst03mQAIzYxseLeffAFX+O2xvcCP9w4EDBv/T2SQSDPBruQG0JbHtMOOGGzkGBxsM2BKI0PIs2SCxLy1xw5k3BYcZDNgM24CBfBCvX84nH5P88c3G3uB4+ubPDH/Y5OXb2w8++FGBWwt2cIBE9aNgFIyCUTAK0AAAzNxeJYG18ukAAAAASUVORK5CYII=","orcid":"","institution":"Xiaomi EV","correspondingAuthor":true,"prefix":"","firstName":"Xiaoshuai","middleName":"","lastName":"Hao","suffix":""},{"id":569015867,"identity":"b5b72e14-68b1-4900-b6b0-273e5b3873db","order_by":1,"name":"Lei Zhou","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Lei","middleName":"","lastName":"Zhou","suffix":""},{"id":569015868,"identity":"dc935ceb-e188-41c0-b53d-9ef6c22339bd","order_by":2,"name":"Zhijian Huang","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Zhijian","middleName":"","lastName":"Huang","suffix":""},{"id":569015869,"identity":"951645c6-15c0-4b4f-aede-8fc47cf76276","order_by":3,"name":"Zhiwen Hou","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Zhiwen","middleName":"","lastName":"Hou","suffix":""},{"id":569015870,"identity":"c1689c7b-64b0-439b-b67f-a3e99ddbe434","order_by":4,"name":"Yingbo Tang","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Yingbo","middleName":"","lastName":"Tang","suffix":""},{"id":569015871,"identity":"1ae8e0ab-f159-4a2e-830f-ad63abdc5873","order_by":5,"name":"Lingfeng Zhang","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Lingfeng","middleName":"","lastName":"Zhang","suffix":""},{"id":569015872,"identity":"b8649610-760f-4172-ac3a-d770fdd49232","order_by":6,"name":"Guang Li","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Guang","middleName":"","lastName":"Li","suffix":""},{"id":569015873,"identity":"2ef54289-5215-46fc-bd5b-3cfa19647740","order_by":7,"name":"Zheng Lu","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Zheng","middleName":"","lastName":"Lu","suffix":""},{"id":569015874,"identity":"205e0003-c2a7-40be-b5c5-12077cfb3a2b","order_by":8,"name":"Shuhuai Ren","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Shuhuai","middleName":"","lastName":"Ren","suffix":""},{"id":569015875,"identity":"57b0f216-fc46-4ce4-9609-07c9e1db1448","order_by":9,"name":"Xianhui Meng","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Xianhui","middleName":"","lastName":"Meng","suffix":""},{"id":569015876,"identity":"18c08206-bb43-44ae-863c-093e24feec70","order_by":10,"name":"Yuchen Zhang","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Yuchen","middleName":"","lastName":"Zhang","suffix":""},{"id":569015877,"identity":"b7835931-ed89-466a-a147-c7c08f537c10","order_by":11,"name":"Jing Wu","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Jing","middleName":"","lastName":"Wu","suffix":""},{"id":569015878,"identity":"b739bcd7-405f-4bd0-b3a0-33e3467959e7","order_by":12,"name":"Jinghui Lu","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Jinghui","middleName":"","lastName":"Lu","suffix":""},{"id":569015879,"identity":"8f3b379c-6961-4518-afe0-f03dd0e1a2a8","order_by":13,"name":"Chenxu Dang","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Chenxu","middleName":"","lastName":"Dang","suffix":""},{"id":569015880,"identity":"6880c0b0-b119-4ac4-adb5-dc703e06b3bb","order_by":14,"name":"Jiayi Guan","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Jiayi","middleName":"","lastName":"Guan","suffix":""},{"id":569015881,"identity":"86d8a194-1dd3-4063-ab58-ad31615d5512","order_by":15,"name":"Jianhua Wu","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Jianhua","middleName":"","lastName":"Wu","suffix":""},{"id":569015882,"identity":"7098c9e4-373f-4fa8-b9f6-08ccffec7049","order_by":16,"name":"Zhiyi Hou","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Zhiyi","middleName":"","lastName":"Hou","suffix":""},{"id":569015883,"identity":"ca0c68c3-2f1d-426d-aecc-9b12ef5751d0","order_by":17,"name":"Hanbing Li","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Hanbing","middleName":"","lastName":"Li","suffix":""},{"id":569015884,"identity":"7f292318-4433-421d-b20c-d24ca1fab05f","order_by":18,"name":"Shumeng Xia","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Shumeng","middleName":"","lastName":"Xia","suffix":""},{"id":569015885,"identity":"13e2097f-737a-4130-be07-a41bd9b1390d","order_by":19,"name":"Mingliang Zhou","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Mingliang","middleName":"","lastName":"Zhou","suffix":""},{"id":569015886,"identity":"2ce4d678-1df8-4c1b-8f1c-44265141ec99","order_by":20,"name":"Yinan Zheng","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Yinan","middleName":"","lastName":"Zheng","suffix":""},{"id":569015887,"identity":"fa95e756-2685-4c4a-9261-169745f67a18","order_by":21,"name":"Zihao Yue","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Zihao","middleName":"","lastName":"Yue","suffix":""},{"id":569015888,"identity":"19ef9427-bd01-49f8-b7cd-c2dd3208604d","order_by":22,"name":"Shuhao Gu","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Shuhao","middleName":"","lastName":"Gu","suffix":""},{"id":569015889,"identity":"4b6aeb9c-22dc-4f6d-b2e8-c1264aadbad5","order_by":23,"name":"Hao Tian","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Hao","middleName":"","lastName":"Tian","suffix":""},{"id":569015890,"identity":"d6588257-93f2-4fe4-a287-7548a0875161","order_by":24,"name":"Yuannan Shen","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Yuannan","middleName":"","lastName":"Shen","suffix":""},{"id":569015891,"identity":"4fef9ac8-7f6e-41e3-9181-1850a157ed57","order_by":25,"name":"Jianwei Cui","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Jianwei","middleName":"","lastName":"Cui","suffix":""},{"id":569015892,"identity":"b0b32eb4-4f32-4ee0-8a2d-da36f1ef567b","order_by":26,"name":"Wen Zhang","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Wen","middleName":"","lastName":"Zhang","suffix":""},{"id":569015893,"identity":"0a71e286-85c5-44ca-9c7e-a9034b9c3951","order_by":27,"name":"Shaoqing Xu","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Shaoqing","middleName":"","lastName":"Xu","suffix":""},{"id":569015894,"identity":"f3444963-1fc2-420f-9978-45d8657ae9a6","order_by":28,"name":"Bing Wang","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Bing","middleName":"","lastName":"Wang","suffix":""},{"id":569015895,"identity":"269ad1f7-a741-4f5d-9630-c152f63ff7f4","order_by":29,"name":"Haiyang Sun","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Haiyang","middleName":"","lastName":"Sun","suffix":""},{"id":569015896,"identity":"fe48735a-0336-4ac6-be17-3353888e19a5","order_by":30,"name":"Zeyu Zhu","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Zeyu","middleName":"","lastName":"Zhu","suffix":""},{"id":569015897,"identity":"ea725a7f-f6c3-416c-8571-20af1a4e347c","order_by":31,"name":"Yuncheng Jiang","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Yuncheng","middleName":"","lastName":"Jiang","suffix":""},{"id":569015898,"identity":"eaab20ea-2e5f-4e99-b1c4-894e57739cbe","order_by":32,"name":"Zibin Guo","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Zibin","middleName":"","lastName":"Guo","suffix":""},{"id":569015899,"identity":"d8ee2e2f-7b6f-458e-bde8-a53dc8610023","order_by":33,"name":"Chuhong Gong","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Chuhong","middleName":"","lastName":"Gong","suffix":""},{"id":569015900,"identity":"eaca773b-d840-4a35-86c7-ec6a32a39355","order_by":34,"name":"Chaofan Zhang","email":"","orcid":"","institution":"Institute of Automation, Chinese Academy of Sciences","correspondingAuthor":false,"prefix":"","firstName":"Chaofan","middleName":"","lastName":"Zhang","suffix":""},{"id":569015901,"identity":"7b8d0646-f795-4b33-91e8-5022dbfc0660","order_by":35,"name":"Wenbo Ding","email":"","orcid":"https://orcid.org/0000-0002-0597-4512","institution":"Tsinghua University","correspondingAuthor":false,"prefix":"","firstName":"Wenbo","middleName":"","lastName":"Ding","suffix":""},{"id":569015902,"identity":"5b087728-cff4-45f3-b6b6-6a0c1d808c4e","order_by":36,"name":"Kun Ma","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Kun","middleName":"","lastName":"Ma","suffix":""},{"id":569015903,"identity":"c82d9314-958b-4792-b7dd-3b875bcf4605","order_by":37,"name":"Guang Chen","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Guang","middleName":"","lastName":"Chen","suffix":""},{"id":569015904,"identity":"1c0d0bb0-4580-455b-8fc8-80dc514c9c75","order_by":38,"name":"Rui Cai","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Rui","middleName":"","lastName":"Cai","suffix":""},{"id":569015905,"identity":"78967942-3682-4934-bc5a-d77b95dfe9e0","order_by":39,"name":"Diyun Xiang","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Diyun","middleName":"","lastName":"Xiang","suffix":""},{"id":569015906,"identity":"667a600a-a221-4045-a3d0-31b8c2a9a9b1","order_by":40,"name":"Heng Qu","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Heng","middleName":"","lastName":"Qu","suffix":""},{"id":569015907,"identity":"8e28d302-7e75-4e04-940f-a8c95009a9f1","order_by":41,"name":"Fuli Luo","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Fuli","middleName":"","lastName":"Luo","suffix":""},{"id":569015908,"identity":"381a8375-2227-46ba-a9aa-254c83ae46b9","order_by":42,"name":"Hangjun Ye","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Hangjun","middleName":"","lastName":"Ye","suffix":""},{"id":569015909,"identity":"af288dcb-6f67-4fee-9995-9a0821d0e7b0","order_by":43,"name":"Long Chen","email":"","orcid":"","institution":"Xiaomi EV","correspondingAuthor":false,"prefix":"","firstName":"Long","middleName":"","lastName":"Chen","suffix":""}],"badges":[],"createdAt":"2026-01-01 06:00:16","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8493355/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8493355/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":101943443,"identity":"026e327b-e16c-457a-8b1e-f0c5c08d07bb","added_by":"auto","created_at":"2026-02-05 09:41:58","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":17324171,"visible":true,"origin":"","legend":"Article File","description":"","filename":"NatureMiMoEmbodied.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8493355/v1_covered_f3b58901-7692-4249-9bb8-de3bc7fc5b63.pdf"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"MiMo-Embodied: X-Embodied Foundation Model","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Cross-Embodied Foundation Models, Embodied AI, Autonomous Driving","lastPublishedDoi":"10.21203/rs.3.rs-8493355/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8493355/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"MiMo-Embodied is a groundbreaking cross-embodied foundation model that integrates indoor robotics and outdoor autonomous driving in one model, effectively addressing the historical ''domain gap'' that has siloed Embodied AI and driving systems. The model achieves state-of-the-art performance across 29 benchmarks, comprising 17 tasks in Embodied AI (including task planning, affordance prediction and spatial understanding) and 12 in autonomous driving (spanning environmental perception, status prediction, and driving planning). Central to these gains is a multi-stage training strategy that couples curated general and domain-specific data with supervised alignment, chain-of-thought reasoning supervision, and reinforcement learning, yielding robust positive transfer and mutual reinforcement between embodiments. The model consistently outperforms specialized, open-source, and closed-source counterparts while retaining general visual semantic understanding. These results indicate that a single vision-language foundation model can acquire cohesive physical intelligence across diverse embodiments and environments, suggesting that carefully staged, multimodal training is sufficient to unlock cross-task generalization without architectural specialization. This work offers a scalable framework for developing unified embodied systems. Code and models are available at https://github.com/XiaomiMiMo/MiMo-Embodied.","manuscriptTitle":"MiMo-Embodied: X-Embodied Foundation Model","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-02-05 03:35:54","doi":"10.21203/rs.3.rs-8493355/v1","editorialEvents":[],"status":"published","journal":{"display":false,"email":"[email protected]","identity":"nature","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"nature","sideBox":"Learn more about [Nature](http://www.nature.com/nature/)","snPcode":"","submissionUrl":"","title":"Nature","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"1ee76aaa-db6d-46dd-9738-219431a5b70f","owner":[],"postedDate":"February 5th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":60544419,"name":"Biological sciences/Biological techniques/Software"},{"id":60544420,"name":"Scientific community and society/Scientific community"}],"tags":[],"updatedAt":"2026-04-23T13:10:46+00:00","versionOfRecord":[],"versionCreatedAt":"2026-02-05 03:35:54","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8493355","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8493355","identity":"rs-8493355","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}