{"id":47548,"date":"2025-12-19T14:50:03","date_gmt":"2025-12-19T06:50:03","guid":{"rendered":"https:\/\/www.1ai.net\/?p=47548"},"modified":"2025-12-19T14:50:03","modified_gmt":"2025-12-19T06:50:03","slug":"%e7%be%8e%e5%9b%a2%e5%bc%80%e6%ba%90%e8%99%9a%e6%8b%9f%e4%ba%ba%e8%a7%86%e9%a2%91%e7%94%9f%e6%88%90%e6%a8%a1%e5%9e%8b-longcat-video-avatar%ef%bc%9a%e5%8f%b7%e7%a7%b0%e4%b8%8d%e8%af%b4","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/47548.html","title":{"rendered":"Open-source virtual human video generation model LongCat-Video-Avatar: It's all human when it's called \"no talking.\""},"content":{"rendered":"<p>On December 19, according to a tweet from the \"LongCat\" public<a href=\"https:\/\/www.1ai.net\/en\/tag\/%e7%be%8e%e5%9b%a2\" title=\"[Sees articles with [American] labels]\" target=\"_blank\" >Meituan (Japanese company)<\/a> The LongCat team officially released and<a href=\"https:\/\/www.1ai.net\/en\/tag\/%e5%bc%80%e6%ba%90\" title=\"[View articles tagged with [open source]]\" target=\"_blank\" >Open Source<\/a> SOTA LEVEL<a href=\"https:\/\/www.1ai.net\/en\/tag\/%e8%99%9a%e6%8b%9f%e4%ba%ba\" title=\"[Sees articles with [virtual] labels]\" target=\"_blank\" >Virtual Human<\/a><a href=\"https:\/\/www.1ai.net\/en\/tag\/%e8%a7%86%e9%a2%91%e7%94%9f%e6%88%90%e6%a8%a1%e5%9e%8b\" title=\"_Other Organiser\" target=\"_blank\" >Video Generation Model<\/a> \u2014 LongCat-Video-Avatar\u3002<\/p>\n<p>Based on the LongCat-Video base, the model is based on the continuation of the core design of a \u201cone model for multitasking\u201d, with originals supporting core functions such as Audio-Text-to-Video, Audio-Text-Image-to-Video, and video continuation, while at the same time fully upgrading the bottom structure to achieve a breakthrough in the three dimensions of operational integrity, long video stability and identity consistency\u3002<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-47549\" title=\"16c673a9j00t7i89y001gd000ucnm\" src=\"https:\/\/www.1ai.net\/wp-content\/uploads\/2025\/12\/16c673a9j00t7i89y001gd000u000cnm.jpg\" alt=\"16c673a9j00t7i89y001gd000ucnm\" width=\"1080\" height=\"455\" \/><\/p>\n<p>According to the official presentation, the model has the following technical highlights\u3002<\/p>\n<p><strong>\"Goodbye, stiffness, for life\":<\/strong>It's not just the mouth, it's the mouth<strong>Synchronize eye, face and body movements<\/strong>To achieve a rich emotional expression\u3002<\/p>\n<p><strong>Even when you don't talk, you look like a man<\/strong>The group uses Disentangled Unconditional Guidance to train models<strong>I understand \"quiet\" doesn't mean \"dead machine.\"<\/strong>I don't know. When you talk, you'll be human<strong>Blinking, repositioning, relaxing<\/strong>.<\/p>\n<p>LongCat-Video-Avatar was described as the first \u201call-power player\u201d to support three modes of production of text, pictures and videos, and the virtual person had \u201creal life power\u201d\u3002<\/p>\n<p>Quantitative assessments on authoritative public data sets such as HDTF, CelebV-HQ, EEMTD and EvalTalker indicate that LongCat-Video-Avatar has achieved SOTA lead on a number of core indicators\u3002<\/p>\n<p><strong>1 Project address:<\/strong><\/p>\n<ul>\n<li><strong>GitHub:<\/strong><strong>\u00a0<\/strong>https:\/\/github.com\/meituan-longcat\/LongCat-Video<\/li>\n<li><strong>Hugging Face:<\/strong><strong>\u00a0<\/strong>https:\/\/huggingface.co\/meituan-longcat\/LongCat-Video-Avatar<\/li>\n<li><strong>Project:<\/strong><strong>\u00a0<\/strong>https:\/\/meigen-ai.github.io\/LongCat-Video-Avatar\/<\/li>\n<\/ul>","protected":false},"excerpt":{"rendered":"<p>On December 19, according to a tweet from the LongCat public, the company LongCat team officially released and opened the SOTA-class virtual human video generation model, LongCat-Video-Avatar. Based on the LongCat-Video base, the model is based on the continuation of the core design of a \u201cone model for multitasking\u201d, with originals supporting core functions such as Audio-Text-to-Video, Audio-Text-Image-to-Video, and video continuation, while at the same time fully upgrading the bottom structure to achieve a breakthrough in the three dimensions of operational integrity, long video stability and identity consistency. According to the official presentation, the model has the following technical highlights\u3002<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[146],"tags":[219,4871,1900,460],"collection":[],"class_list":["post-47548","post","type-post","status-publish","format-standard","hentry","category-news","tag-219","tag-4871","tag-1900","tag-460"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/47548","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/comments?post=47548"}],"version-history":[{"count":0,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/47548\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/media?parent=47548"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/categories?post=47548"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/tags?post=47548"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/collection?post=47548"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}