{"id":544124,"date":"2026-07-16T09:59:12","date_gmt":"2026-07-16T09:59:12","guid":{"rendered":"https:\/\/www.newsbeep.com\/il\/544124\/"},"modified":"2026-07-16T09:59:12","modified_gmt":"2026-07-16T09:59:12","slug":"gemma-4-gets-a-stealth-update-that-fixes-tool-calling-bugs-and-truncated-responses-under-the-same-name","status":"publish","type":"post","link":"https:\/\/www.newsbeep.com\/il\/544124\/","title":{"rendered":"Gemma 4 gets a stealth update that fixes tool calling bugs and truncated responses under the same name"},"content":{"rendered":"<p>\t\t\t<a href=\"https:\/\/the-decoder.com\/author\/jonathan-kemper\/\" class=\"w-6 h-6 relative rounded-full overflow-hidden shrink-0 bg-gray-400 block   exclude-from-pdf\" title=\"View all posts by Jonathan Kemper\" rel=\"nofollow noopener\" target=\"_blank\"><br \/>\n\t\t\t\t<img loading=\"lazy\" alt=\"Jonathan Kemper\" src=\"https:\/\/www.newsbeep.com\/il\/wp-content\/uploads\/2026\/07\/avatar_jonathan_kemper_compressed-375x375.jpeg\"  class=\"avatar avatar-24 photo w-6 h-6 left-0 top-0 absolute\" height=\"24\" width=\"24\" decoding=\"async\"\/>\t\t\t<\/a><\/p>\n<p>Google shipped an update to its open AI model Gemma 4 that speeds up performance on Nvidia Hopper GPUs, fixes tool calling bugs, and addresses problems with truncated responses. Turning on Flash Attention 4 boosts the speed at which the model processes incoming prompts by 25 to 70 percent, according to Google. Time to first token drops by up to 31 percent. Google also fixed bugs in tool calling, the feature that lets the model trigger external tools on its own.<\/p>\n<p><img fetchpriority=\"high\" decoding=\"async\" class=\"size-full wp-image-58829\" src=\"https:\/\/www.newsbeep.com\/il\/wp-content\/uploads\/2026\/07\/Google-Gemma-4-31B-net-improvements-scaled-1.jpeg\" alt=\"Gemma 4 31B: BFCL 74,2 %(+0,4), TB2 25,8 %(+4,5), Tau2 Retail 77,6 %(+3,1), Airline 84 %(+2,0), Telecom 62,7 %(+10,1).\" width=\"2560\" height=\"1440\"\/>Gemma 4 31B improves its agentic reasoning and tool-calling performance across all tested scenarios, with the highest increase of 10.1% in the telecommunications use case.<\/p>\n<p>Google says it also cut down on cases where the model would cut answers short or return incomplete responses. For image processing, users can manually raise the &#8220;max_soft_tokens&#8221; parameter from 280 to 1,120 to get sharper OCR results and support resolutions up to 2.51 megapixels. Google put up an <a href=\"https:\/\/huggingface.co\/spaces\/google\/gemma4_vision_token_budget\" target=\"_blank\" rel=\"noopener noreferrer nofollow\">interactive configurator on Hugging Face<\/a> for that. The published benchmarks only compare the 31B and E4B variants against their predecessors, but the <a href=\"https:\/\/huggingface.co\/collections\/google\/gemma-4\" target=\"_blank\" rel=\"noopener noreferrer nofollow\">Hugging Face repository<\/a> shows that all parameter sizes in this model generation got updated, including <a href=\"https:\/\/the-decoder.com\/google-deepminds-gemma-4-12b-squeezes-multimodal-ai-onto-a-laptop-with-just-16-gb-of-ram\/\" target=\"_blank\" rel=\"noopener noreferrer nofollow\">the newest 12B release<\/a>. The community has pushed back on Google shipping the update under the same <a href=\"https:\/\/the-decoder.com\/googles-gemma-4-is-now-available-with-apache-2-0-licensing-for-the-first-time\/\" target=\"_blank\" rel=\"noopener noreferrer nofollow\">&#8220;Gemma 4&#8221; name<\/a> instead of tagging it as a separate version like &#8220;Gemma 4.1.&#8221;<\/p>\n<p>\t\t\t\tAI News Without the Hype \u2013 Curated by Humans<\/p>\n<p>\n\t\t\t\t\tSubscribe to THE DECODER for ad-free reading, a weekly AI newsletter, our exclusive &#8220;AI Radar&#8221; frontier report six times a year, full archive access, and access to our comment section.\t\t\t\t<\/p>\n<p>\t\t\t\t<a href=\"https:\/\/the-decoder.com\/subscription\/\" class=\"inline-block text-white bg-(--heise-primary) mt-3 hover:bg-blue-800 focus:ring-4 focus:outline-none focus:ring-blue-300 font-medium rounded-sm w-full sm:w-auto  pl-3 pr-3 py-2.5 text-center newsletter-submit-button hover:no-underline\" rel=\"nofollow noopener\" target=\"_blank\"><br \/>\n\t\t\t\t\tSubscribe now\t\t\t\t<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"Google shipped an update to its open AI model Gemma 4 that speeds up performance on Nvidia Hopper&hellip;\n","protected":false},"author":2,"featured_media":544125,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[6],"tags":[138571,133,85,46,125],"class_list":["post-544124","post","type-post","status-publish","format-standard","has-post-thumbnail","category-technology","tag-gemma","tag-google","tag-il","tag-israel","tag-technology"],"_links":{"self":[{"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/posts\/544124","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/comments?post=544124"}],"version-history":[{"count":0,"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/posts\/544124\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/media\/544125"}],"wp:attachment":[{"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/media?parent=544124"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/categories?post=544124"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.newsbeep.com\/il\/wp-json\/wp\/v2\/tags?post=544124"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}