{"id":19448,"date":"2024-09-06T09:38:33","date_gmt":"2024-09-06T01:38:33","guid":{"rendered":"https:\/\/www.1ai.net\/?p=19448"},"modified":"2024-09-06T09:38:33","modified_gmt":"2024-09-06T01:38:33","slug":"%e9%9d%a2%e5%a3%81%e6%99%ba%e8%83%bd%e5%8f%91%e5%b8%83%e5%b0%8f%e9%92%a2%e7%82%aeminicpm-3-0%e7%ab%af%e4%be%a7%e6%a8%a1%e5%9e%8b%ef%bc%9a2gb%e5%86%85%e5%ad%98%e5%8d%b3%e5%8f%af%e8%bf%90%e8%a1%8c","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/19448.html","title":{"rendered":"Mianbi Intelligent released the MiniCPM 3.0 client-side model: it can run with 2GB of memory and its performance exceeds GPT-3.5"},"content":{"rendered":"<p><a href=\"https:\/\/www.1ai.net\/en\/tag\/%e9%9d%a2%e5%a3%81%e6%99%ba%e8%83%bd\" title=\"[View articles tagged with [face smart]]\" target=\"_blank\" >Wall-facing intelligence<\/a>The official WeChat account published a blog post yesterday (September 5) announcing the launch of<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> MiniCPM3-4B <a href=\"https:\/\/www.1ai.net\/en\/tag\/ai%e6%a8%a1%e5%9e%8b\" title=\"[View articles tagged with [AI models]]\" target=\"_blank\" >AI Models<\/a>, claiming that &quot;the time for on-device ChatGPT has arrived&quot;. This is an AI model with excellent performance, capable of running on devices with only 2GB of memory, heralding a new era of on-device AI experience.<\/p>\n<p>The MiniCPM3.0 model has 4B parameters and outperforms GPT-3.5 in performance, enabling AI services on mobile devices that are comparable to GPT-3.5. This allows users to enjoy fast, secure, and feature-rich local AI services without relying on cloud processing, while ensuring a smoother and more private intelligent interaction experience.<\/p>\n<p><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-19449\" title=\"e060ec81j00sjdb64000xd000e300hzm\" src=\"https:\/\/www.1ai.net\/wp-content\/uploads\/2024\/09\/e060ec81j00sjdb64000xd000e300hzm.jpg\" alt=\"e060ec81j00sjdb64000xd000e300hzm\" width=\"507\" height=\"647\" \/><\/p>\n<p>MiniCPM3.0 features include unlimited length text processing capability, the strongest end-side Function Calling performance, and powerful RAG (Retrieval Enhanced Generation) capabilities. These features give MiniCPM3.0 significant advantages in processing long texts, executing structured instructions, and providing accurate information.<\/p>\n<p>Li Dahai, CEO of Mianbi Intelligence, said that the release of MiniCPM3.0 is a big step forward for edge AI technology. It will promote the implementation of large AI models in practical applications and provide users with a richer and more convenient intelligent experience.<\/p>\n<p>It is expected that in the near future, more products equipped with the MiniCPM3.0 model will be launched on the market, further promoting the popularization and application of AI technology.<\/p>\n<p><strong>MiniCPM3.0 open source address:<\/strong><\/p>\n<p>https:\/\/github.com\/OpenBMB\/MiniCPM<\/p>\n<p>https:\/\/huggingface.co\/openbmb\/MiniCPM3-4B<\/p>","protected":false},"excerpt":{"rendered":"<p>In a blog post published yesterday (September 5), the official public number of Facade Intelligence announced the launch of the open source MiniCPM3-4B AI model, claiming that \"the time for end-side ChatGPT has arrived\". This is a high-performance AI model that can run on devices with as little as 2GB of RAM, heralding a new era of end-side AI experiences. With 4B parameters, the MiniCPM3.0 model outperforms GPT-3.5 and is capable of delivering AI services on mobile devices at a level comparable to GPT-3.5. This allows users to enjoy fast, secure and feature-rich local AI services without relying on cloud processing, while ensuring a smoother and more private intelligent interaction experience. MiniCPM3.0 features include unlimited long text processing capability, the strongest end-side F<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[146],"tags":[167,219,2184],"collection":[],"class_list":["post-19448","post","type-post","status-publish","format-standard","hentry","category-news","tag-ai","tag-219","tag-2184"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/19448","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=19448"}],"version-history":[{"count":0,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/19448\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/media?parent=19448"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/categories?post=19448"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/tags?post=19448"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/collection?post=19448"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}