{"id":50228,"date":"2026-02-17T22:10:00","date_gmt":"2026-02-17T14:10:00","guid":{"rendered":"https:\/\/www.1ai.net\/?p=50228"},"modified":"2026-02-17T22:10:00","modified_gmt":"2026-02-17T14:10:00","slug":"%e9%98%bf%e9%87%8c%e5%bc%80%e6%ba%90%e6%97%97%e8%88%b0qwen3-5-%e5%8f%91%e5%b8%83%ef%bc%8c%e7%99%bb%e9%a1%b6%e5%85%a8%e7%90%83%e6%9c%80%e5%bc%ba%e5%bc%80%e6%ba%90%e6%a8%a1%e5%9e%8b","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/50228.html","title":{"rendered":"Ali Open Flag Qwen3.5 released, top of the world's strongest open source model"},"content":{"rendered":"

February 17th.Ali<\/a>Qwen3.5-Plus and Qwen3.5-397B-A17B new models are down on the Chat.qwen.ai page\u3002<\/p>\n

\"1a9feff2j00talwna000td000o000dop\"<\/p>\n

1AI learned from the official page that the Qwen3.5 Plus location is\u00a0Qwen 3.5 Recent Large Language Model Series<\/strong>Qwen3.5-397B-A17B\u00a0Qwen3.5 Open Source Series flagship language model<\/strong>I don't know. Two modelsBoth text and polymorphic tasks are supported<\/strong>.<\/p>\n

According to Aliyun, Qwen3.5 achieved a complete overhaul of the bottom model structure, of which Qwen3.5-Plus, with a total parameter of 39.7 billion, activates only 17 billionQwen3-Max model with more than trillion parameters<\/strong>60%, WITH A SIGNIFICANT INCREASE IN REASONING EFFICIENCY, WITH THE LARGEST AMOUNT OF REASONING THROUGHPUT INCREASING TO 19 TIMES\u3002<\/p>\n

Qwen3.5 scored 87.8 in the MMLU-Pro cognitive assessmentBEYOND GPT-5.2<\/strong>; 88.4 POINTS IN THE DOCTORAL CHALLENGE GPQA TESTHigher than Claude 4.5<\/strong>; following IFBench at 76.5 minutesRefresh all model records<\/strong>;in the generic Agent assessment of BFCL-V4, search for Agent assessment of Brownsecomp, thousands of questions about 3.5 performanceBeyond Gemini 3 Pro<\/strong>.<\/p>\n

Qwen3.5-397B-A17B performed well in the full range of benchmarking assessments, such as reasoning, programming, intelligent body abilities and multimodular understanding, and helped developers and enterprises to significantly increase productivity. Using an innovative hybrid structure, the model combines the linear Delta Networks with the rare hybrid (MoE) to achieve excellent reasoning efficiency:Parameters amount to 3970 billion, and only 17 billion parameters are activated at each forward transmission<\/strong>Optimizing speed and cost while maintaining capacity. At the same time, language and dialect support was expanded from 119 to 201, providing wider availability and improved support to users worldwide\u3002<\/p>\n

Qwen3.5 Advances pre-training on three dimensions of competence, efficiency and interoperability:<\/p>\n