{"id":22459,"date":"2025-04-29T12:17:56","date_gmt":"2025-04-29T12:17:56","guid":{"rendered":"https:\/\/metaverseplanet.net\/blog\/?p=22459"},"modified":"2026-01-05T09:16:12","modified_gmt":"2026-01-05T09:16:12","slug":"alibaba-unveils-new-ai-model-family-qwen3","status":"publish","type":"post","link":"https:\/\/metaverseplanet.net\/blog\/alibaba-unveils-new-ai-model-family-qwen3\/","title":{"rendered":"Alibaba Unveils New AI Model Family Qwen3"},"content":{"rendered":"\n

Chinese e-commerce giant Alibaba<\/strong> has officially introduced its new artificial intelligence model family<\/strong>, called Qwen3<\/strong>. According to performance tests, Qwen3<\/strong> has outperformed models from OpenAI<\/strong> and Google<\/strong> in several areas. So, what does the Qwen3 family<\/strong> promise?<\/p>\n\n\n\n

\n\n\n\n

Qwen3: A Hybrid AI Model Built for Reasoning<\/strong><\/h2>\n\n\n\n
\"\"<\/figure>\n\n\n\n

After years of AI development, Alibaba<\/strong> showcased Qwen3<\/strong> at a recent event, claiming it can reason more effectively than its top competitors in certain scenarios. Designed as a hybrid AI model<\/strong>, Qwen3<\/strong> doesn\u2019t waste extra resources on simple tasks but demonstrates advanced reasoning<\/strong> capabilities during complex operations.<\/p>\n\n\n\n

The Qwen3 family<\/strong> features a range of models with parameter sizes from 0.6 billion to 235 billion<\/strong>, offering scalable performance. This design allows the models to be efficient for lightweight tasks while still being capable of handling more demanding ones with precision.<\/p>\n\n\n\n

\n\n\n\n

Qwen-3-235B-A22B<\/strong>: The Most Powerful Model Yet<\/h2>\n\n\n\n
\"\"<\/figure>\n\n\n\n

According to Alibaba<\/strong>, the most powerful model in the lineup is Qwen-3-235B-A22B<\/strong>, which focuses heavily on programming-related tasks<\/strong>. It has already outperformed OpenAI\u2019s o3-mini<\/strong>, Google\u2019s Gemini 2.5 Pro<\/strong>, and Deepseek\u2019s R1<\/strong> in benchmark tests.<\/p>\n\n\n\n

However, this particular model is not yet available for public use. Still, several other models from the Qwen3 family<\/strong> have already been released on Hugging Face<\/strong> and GitHub<\/strong>. While these models are currently in demo mode<\/strong>, they are open-source<\/strong> and accessible to developers and researchers.<\/p>\n\n\n\n

\n\n\n\n

Multilingual Support and Massive Training Dataset<\/strong><\/h2>\n\n\n\n\n\n

The Qwen3 models<\/strong> have been trained on a vast dataset of 36 trillion tokens<\/strong> and offer support for 119 languages<\/strong>. If they continue to deliver strong results, these models could attract considerable attention from both developers and enterprises around the world.<\/p>\n\n\n\n

\n\n\n\n

Alibaba\u2019s Entry into the Global AI Race<\/strong><\/h2>\n\n\n\n
\"\"<\/figure>\n\n\n\n

With the introduction of Qwen3<\/strong>, Alibaba<\/strong> is positioning itself as a serious contender in the AI development landscape<\/strong>, challenging Western tech giants like OpenAI<\/strong> and Google<\/a><\/em><\/strong>. The launch signals a broader ambition to make cutting-edge AI tools<\/strong> more accessible and efficient for global use.<\/p>\n\n\n\n

You Might Also Like;<\/h3>\n\n\n
    <\/ul>","protected":false},"excerpt":{"rendered":"

    Chinese e-commerce giant Alibaba has officially introduced its new artificial intelligence model family, called Qwen3. According to performance tests, Qwen3 has outperformed models from OpenAI and Google in several areas. So, what does the Qwen3 family promise? Qwen3: A Hybrid AI Model Built for Reasoning After years of AI development, Alibaba showcased Qwen3 at a …<\/p>\n","protected":false},"author":1,"featured_media":22462,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"googlesitekit_rrm_CAown96uCw:productID":"","footnotes":""},"categories":[332],"tags":[335,210],"class_list":["post-22459","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ai-information","tag-ai-news","tag-ai-tools-news"],"amp_enabled":true,"_links":{"self":[{"href":"https:\/\/metaverseplanet.net\/blog\/wp-json\/wp\/v2\/posts\/22459","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/metaverseplanet.net\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/metaverseplanet.net\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/metaverseplanet.net\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/metaverseplanet.net\/blog\/wp-json\/wp\/v2\/comments?post=22459"}],"version-history":[{"count":0,"href":"https:\/\/metaverseplanet.net\/blog\/wp-json\/wp\/v2\/posts\/22459\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/metaverseplanet.net\/blog\/wp-json\/wp\/v2\/media\/22462"}],"wp:attachment":[{"href":"https:\/\/metaverseplanet.net\/blog\/wp-json\/wp\/v2\/media?parent=22459"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/metaverseplanet.net\/blog\/wp-json\/wp\/v2\/categories?post=22459"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/metaverseplanet.net\/blog\/wp-json\/wp\/v2\/tags?post=22459"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}