Qwen Office Launches 10 Models Including Qwen4-Max, Kimi-K3.1, and GLM-5.3
On November 16, Qwen Office launched 10 models, including the newly released Qwen4-Max as well as third-party large models from Kimi, GLM, DeepSeek, MiniMax, Step, and others; Personal edition and Enterprise Standard edition users can choose freely. Previously, Qwen Office offered only three tiers of models: Standard, Advanced, and Qwen3.8-Max; this expansion is intended to meet users' more personalized Agent scenario needs. Enterprise administrators can also configure model permissions by department, user group, or member.
Why read this
AI score 58
The original article lists the task types and cost differences each of the ten models is suited for, so readers can judge how multi-model choice affects Agent use in office scenarios.
Topics
Body · AI translation
Top banner ad space
You are using an old version of IE browser. For the security of your Leiphone account and a better product experience, we strongly recommend using a faster and safer browser.
Model choice freedom! Qianwen Office launches 10 models including Qwen4, Kimi-K3.1, GLM-5.3
Author of this article: Xu Mi
2026-11-16 12:08
Lead: On November 16, Qianwen Office launched 10 models at once, including the just-released Qwen4-Max, as well as multiple third-party large models such as Kimi, GLM, DeepSeek, MiniMax, and Step.
On November 16, Qianwen Office launched 10 models at once, including the just-released Qwen4-Max, as well as multiple third-party large models such as Kimi, GLM, DeepSeek, MiniMax, and Step. Currently, Qianwen Office's model list is open for users to choose from, and personal edition and enterprise standard edition users can freely select models based on their own usage scenarios.
Previously, Qianwen Office offered users three tiers of model choices for model supply: Standard, Advanced, and Qwen3.8-Max, which could meet the needs of most daily office tasks. However, as users' Agent usage scenarios continue to deepen, users' demands for the models behind Agents are becoming increasingly personalized. The core reason is that different models have significant differences in effectiveness, response speed, and cost on tasks of different complexity. To this end, Qianwen Office provides richer model choices on top of the original three tiers, allowing users to freely choose the model that best fits different task requirements and efficiently complete tasks at lower token costs.
For example, in tasks requiring multi-step complex reasoning, users can choose Qwen4-Max. According to the introduction, this model is Alibaba's new-generation foundation large model, which implements a continuous delivery mechanism within the model: it can first generate a usable initial version and continuously expand the content version by version according to user feedback until completion; at the same time, the model also has the ability to perform operations research modeling, solving, and verification around complex goals and constraints, and through autonomous planning and continuous iteration, orchestrates agents to solve more complex tasks, easily completing complex tasks such as website design and cloud resource system scheduling.
When handling tasks such as rule sorting, plan checking, simulation demonstrations, and 3D displays, which require the model to process and complete analysis and judgment work, users can choose models such as GLM-5.3 and DeepSeek-V4-Pro. In daily tasks requiring fast responses, such as copywriting, information extraction, and content organization, users only need to choose lower-priced models such as Qwen3.8-Flash, GLM-5.3-Flash, and DeepSeek-V4.1-Flash.
According to the introduction, Qianwen Office also provides model management permissions for enterprise users. Enterprise administrators can configure model permissions by department, user group, or member, providing the most suitable models for different positions and work of different complexity, making it convenient for enterprises to uniformly manage AI assets.
South Korean consumers fall in love with Double 11! Chinese brands' sales on AliExpress surge 4 times
Alibaba officially releases Qwen4, with advanced agent and programming capabilities, moving toward real ...
Foreign trade also welcomes a good start to Double 11! Alibaba.com's US promotion surges over 40 ...
The OCR 3.0 era has arrived! Wenxin Unlimited OCR 2 open-sourced and released, ultra-long document parsing welcomes "page count freedom"
Double 11 iPhone Duo platform battle, what matters is supply, reporter's actual test: only Apple's official website and Tmall can place orders
Double 11 iPhone Duo supply battle: Why does Apple give more stock to Tmall?
Accio becomes a business accelerator for Double 11 merchants: preheating period shortened by half a month, 15-minute optimization of hit products across the entire network
Latest articles
South Korean consumers fall in love with AliExpress Double 11! Chinese brands' sales surge 4 times
South Korean consumers fall in love with Double 11! Chinese brands' sales on AliExpress surge 4 times
Baidu World 2026 | Making financial AI affordable and more professional, Du Xiaoman's "Xuanyuan" explores a new path for small and medium-sized institutions
Douyin Mall Double 11 Good Goods Festival concludes: How do brands achieve business explosion during major promotions?
Huawei Mate X8 officially announces second-generation Linglong screen, is the foldable phone entering a new era of major upgrades?
Alibaba officially releases Qwen4, with advanced agent and programming capabilities, moving toward real full scenarios
Apple Tencent Cloud NLP Andrew Ng O2O product manager skype Qi An Xin browser world model edge computing
Set now Later
Youlu tracking code