{"id":101206,"date":"2026-07-28T02:29:11","date_gmt":"2026-07-28T02:29:11","guid":{"rendered":"https:\/\/www.europesays.com\/korea\/101206\/"},"modified":"2026-07-28T02:29:11","modified_gmt":"2026-07-28T02:29:11","slug":"kakao-unveiled-four-types-of-lightweight-language-models-slms-of-its-own-artificial-intelligence","status":"publish","type":"post","link":"https:\/\/www.europesays.com\/korea\/101206\/","title":{"rendered":"Kakao unveiled four types of lightweight language models (SLMs) of its own artificial intelligence (.."},"content":{"rendered":"<p>Canana-2 SLM Series<br \/>Global Level Performance<br \/>Optimized for Smartphone Driving  <\/p>\n<p>            <img decoding=\"async\" src=\"https:\/\/www.europesays.com\/korea\/wp-content\/uploads\/2026\/07\/news-p.v1.20260728.c2cf1590e720452fa067e18af573add7_P1.png\" orgwidth=\"700\" orgheight=\"394\" alt=\"Canana lightweight language model. &lt;Kakao&gt;\"\/>                      \uc0ac\uc9c4 \ud655\ub300                Canana lightweight language model.       <\/p>\n<p refid=\"2\">Kakao unveiled four types of lightweight language models (SLMs) of its own artificial intelligence (AI) model &#8220;Kanana&#8221; as open source. Despite its small size, it has implemented global-level performance in both Korean and English.<\/p>\n<p refid=\"3\">On the 28th, Kakao announced that it has unveiled Kanana-2-1.3B-base, Kanana-2-1.3B-instruct, Kanana-2-3B-base, and Kanana-2-3B-instruct, a series of lightweight language models that can run on smartphones, on its huggling face.<\/p>\n<p refid=\"4\">Compared to the latest SOTA open-source models of similar size, the models released this time performed better on most representative benchmarks such as Korean, English, knowledge, mathematics, and code. <\/p>\n<p refid=\"5\">In particular, Kanana-2-1.3B-instruct and Kanana-2-0.9B-instruct showed competitive performance over comparable global models such as Alibaba&#8217;s Qwen and Meta&#8217;s Gemma in actual service utilization capabilities such as conversation, knowledge, code, mathematics, instruction execution, and tool calling.<\/p>\n<p refid=\"6\">In addition, Kakao has improved the efficiency of Korean processing by more than 30% compared to the previous one by applying its own Korean-specific talkizer. Talk Nizer is a key technology that determines the unit that splits sentences into small pieces when AI models process text, and the more Korean-optimized talk Nizer is used, the less units the same sentence can be processed, speeding up the calculation and reducing the cost.<\/p>\n<p refid=\"7\">Kakao has secured both performance and cost competitiveness by applying the Korean-specific talkizer, which was developed last year, to the entire model.<\/p>\n<p refid=\"8\">It is also noteworthy that it introduced a Sliding Window Attention structure that solves the problem of rapidly increasing memory usage when processing long conversations by optimizing it for on-device environments where memory and computational resources are limited, such as smartphones.<\/p>\n<p refid=\"9\">This enables stable performance while reducing memory by up to 72.7% even in conversations up to 32K (32,000 tokens, about 24,000 words) long.<\/p>\n<p refid=\"10\">Kakao will distribute this model under the &#8220;Kanana Open License,&#8221; which allows commercial use. Anyone, such as developers, startups, and research institutes, can use technology without any restrictions.<\/p>\n<p refid=\"11\">&#8220;As we prepare for the agentic AI era, we feel the importance of both large-scale AI in the cloud and on-device lightweight AI,&#8221; said Noh Byung-seok, Kakao&#8217;s unified foundation model performance leader. &#8220;Kakao is increasing its technological competitiveness in both areas, and we hope that this open-source public model will lead more developers and companies to develop new AI services and contribute to revitalizing the domestic AI ecosystem.&#8221;<\/p>\n","protected":false},"excerpt":{"rendered":"Canana-2 SLM SeriesGlobal Level PerformanceOptimized for Smartphone Driving \uc0ac\uc9c4 \ud655\ub300 Canana lightweight language model. Kakao unveiled four types&hellip;\n","protected":false},"author":2,"featured_media":101207,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[622],"tags":[827,15962,52496,52497,2762,1138],"class_list":["post-101206","post","type-post","status-publish","format-standard","has-post-thumbnail","category-kakao","tag-kakao","tag-kanana","tag-language-model","tag-light-language-model","tag-open-source","tag-smartphone"],"_links":{"self":[{"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/posts\/101206","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/comments?post=101206"}],"version-history":[{"count":0,"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/posts\/101206\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/media\/101207"}],"wp:attachment":[{"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/media?parent=101206"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/categories?post=101206"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/tags?post=101206"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}