{"id":5407,"date":"2025-05-21T00:25:43","date_gmt":"2025-05-20T23:25:43","guid":{"rendered":"https:\/\/villpress.com\/?p=5407"},"modified":"2025-05-21T00:25:52","modified_gmt":"2025-05-20T23:25:52","slug":"google-unveils-gemma-3n-ai-model-that-can-run-directly-on-phones","status":"publish","type":"post","link":"https:\/\/villpress.com\/cs\/google-unveils-gemma-3n-ai-model-that-can-run-directly-on-phones\/","title":{"rendered":"Google Unveils Gemma 3n AI Model That Can Run Directly on Phones"},"content":{"rendered":"

At Google I\/O 2025, the tech giant introduced Gemma 3n<\/a><\/strong>, the newest addition to its family of \u201copen\u201d AI models \u2014 and this one is designed to run directly on mobile devices. Starting Tuesday, developers can preview the lightweight yet powerful model, which supports audio, text, images, and video processing.<\/p>\n\n\n\n

According to Google, Gemma 3n<\/strong> is optimized to work \u201csmoothly\u201d on phones, laptops, and tablets\u2014even those with less than\u00a02GB of RAM<\/strong>. This means users can now run advanced AI models offline without relying on cloud computing, making them cheaper and more private<\/span>.<\/p>\n\n\n\n

\n

\u201cGemma 3n shares the same architecture as Gemini Nano, and is engineered for incredible performance,\u201d said Gus Martins, Product Manager for Gemma, during the keynote presentation.<\/p>\n<\/blockquote>\n\n\n\n

Running models locally on-device has become an exciting area in AI development, offering lower latency and enhanced privacy. Gemma 3n is part of Google\u2019s push to make high-performance AI more accessible across a broader range of consumer devices.<\/p>\n\n\n\n

MedGemma: AI for Health Apps<\/h3>\n\n\n\n

Google also unveiled MedGemma<\/strong>, its most capable open model for understanding health-related images and text. Released under its Health AI Developer Foundations<\/strong> program, MedGemma is meant to help developers create smarter, more adaptive healthcare apps.<\/p>\n\n\n\n

\n

\u201cMedGemma works great across a range of image and text applications, so that developers can adapt the models for their own health apps,\u201d said Martins.<\/p>\n<\/blockquote>\n\n\n\n

SignGemma: AI for Sign Language Translation<\/h3>\n\n\n\n

Looking ahead, Google previewed SignGemma<\/strong>, a model designed to translate sign language into spoken-language text \u2014 with a focus on American Sign Language (ASL)<\/strong> and English.<\/p>\n\n\n\n

\n

\u201cSignGemma is the most capable sign language understanding model ever, and we can\u2019t wait for you \u2014 developers and deaf and hard-of-hearing communities \u2014 to take this foundation and build with it,\u201d Martins added.<\/p>\n<\/blockquote>\n\n\n\n

SignGemma is aimed at enabling a new wave of accessibility-focused applications for the deaf and hard-of-hearing community.<\/p>\n\n\n\n

A Note on Licensing<\/h3>\n\n\n\n

While the Gemma models have been downloaded tens of millions of times<\/strong>, some developers have criticized Google\u2019s non-standard licensing terms<\/strong>, which they say may complicate commercial use. Still, the models continue to gain traction due to their flexibility and performance.<\/p>\n\n\n\n

Gemma 3n and its companion models represent Google\u2019s latest attempt to democratize AI development while keeping performance high and hardware requirements low.<\/p>","protected":false},"excerpt":{"rendered":"

At Google I\/O 2025, the tech giant introduced Gemma 3n, the newest addition to its family of \u201copen\u201d AI models \u2014 and this one is designed to run directly on mobile devices. Starting Tuesday, developers can preview the lightweight yet powerful model, which supports audio, text, images, and video processing. According to Google, Gemma 3n […]<\/p>\n","protected":false},"author":1,"featured_media":5408,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_mi_skip_tracking":false,"footnotes":""},"categories":[64],"tags":[351,116],"ppma_author":[331],"class_list":{"0":"post-5407","1":"post","2":"type-post","3":"status-publish","4":"format-standard","5":"has-post-thumbnail","7":"category-ai","8":"tag-gemma-3n","9":"tag-google"},"authors":[{"term_id":331,"user_id":1,"is_guest":0,"slug":"pastakutmanwen","display_name":"Villpress Insider","avatar_url":{"url":"https:\/\/villpress.com\/wp-content\/uploads\/2025\/05\/Logo.png","url2x":"https:\/\/villpress.com\/wp-content\/uploads\/2025\/05\/Logo.png"},"0":null,"1":"","2":"","3":"","4":"","5":"","6":"","7":"","8":""}],"_links":{"self":[{"href":"https:\/\/villpress.com\/cs\/wp-json\/wp\/v2\/posts\/5407","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/villpress.com\/cs\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/villpress.com\/cs\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/villpress.com\/cs\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/villpress.com\/cs\/wp-json\/wp\/v2\/comments?post=5407"}],"version-history":[{"count":1,"href":"https:\/\/villpress.com\/cs\/wp-json\/wp\/v2\/posts\/5407\/revisions"}],"predecessor-version":[{"id":5409,"href":"https:\/\/villpress.com\/cs\/wp-json\/wp\/v2\/posts\/5407\/revisions\/5409"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/villpress.com\/cs\/wp-json\/wp\/v2\/media\/5408"}],"wp:attachment":[{"href":"https:\/\/villpress.com\/cs\/wp-json\/wp\/v2\/media?parent=5407"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/villpress.com\/cs\/wp-json\/wp\/v2\/categories?post=5407"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/villpress.com\/cs\/wp-json\/wp\/v2\/tags?post=5407"},{"taxonomy":"author","embeddable":true,"href":"https:\/\/villpress.com\/cs\/wp-json\/wp\/v2\/ppma_author?post=5407"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}