AI News

Qwen Office Adds 10 New Large Models, Reopens Third-Party Model Selection

WallstreetcnUpdated Nov 16 21:10

AI summary

On November 16, Alibaba's Qwen Office announced the addition of 10 new large models, including the newly released Qwen4-Max, as well as third-party models such as Kimi, GLM, DeepSeek, MiniMax, and Step. Users of both the personal edition and the enterprise standard edition can choose models on their own. This comes just four days after Doubao Work fully opened third-party models to individual users, and Tencent WorkBuddy and Baidu Kuku AI had previously also moved toward a multi-model aggregation approach. Qwen Office had integrated GLM-5.3 and DeepSeek V4 Pro on August 14, but three days later these two external models disappeared from the product, and Alibaba did not publicly explain the reason.

Why read this

AI score 52

After wavering between in-house development and openness, Qwen Office has once again integrated third-party models; readers can use this to observe how competition among office Agents is shifting from model lineup to orchestration capability.

Topics

Tags

Body · AI translation

The multi-model camp for office Agents has finally welcomed Alibaba.

On November 16, Alibaba's Qianwen Office announced the addition of 10 large models, including the newly released Qwen4-Max, as well as third-party models from multiple vendors such as Kimi, GLM, DeepSeek, MiniMax, and Step.

After this update, users of both the Qianwen Office personal edition and the enterprise standard edition can independently choose models, deciding which model to use to complete work based on task complexity, response speed, and Token cost.

This comes only four days after ByteDance's Doubao Work fully opened third-party models to individual users.

At this point, Tencent WorkBuddy, ByteDance Doubao Work, Baidu Kuku AI, and Alibaba Qianwen Office have all moved toward a multi-model aggregation approach, and the model strategies of major companies in office Agents are further converging.

For Qianwen Office, this opening up was somewhat tortuous.

As early as August 14 this year, Qianwen Office announced that it had integrated two external models, Zhipu GLM-5.3 and DeepSeek V4 Pro, officially expanding the range of model choices.

But just three days later, things changed.

On the evening of August 17, Wallstreetcn · All-Weather Technology found that the above two external models had disappeared from the Qianwen Office product side, and the model selector returned to a state centered on different capability tiers and Qwen3.8-Max.

Alibaba did not publicly explain the reason for the sudden removal of the two models.

Now, after about three months, Qianwen Office has once again opened up third-party models, and added 10 at once, further increasing the degree of openness.

This back-and-forth also reflects the multiple-choice question Alibaba faces in the office Agent field.

Alibaba itself is one of the major domestic large-model R&D vendors. From Tongyi Qianwen to Alibaba Cloud Bailian, Alibaba has established a relatively complete model R&D and commercialization system.

For Qwen, Qianwen Office is not only a window to showcase its self-developed model capabilities, but also an important application scenario for undertaking real user needs.

Once third-party models are opened up, some tasks originally assigned to Qwen may shift to other model vendors.

But from the perspective of Qianwen Office's own product competitiveness, opening up external models is also necessary.

Office Agents differ from traditional AI chat products in that the tasks users hand over often involve multiple stages such as information search, data analysis, code writing, tool invocation, and deliverable output. Different tasks have different requirements for a model's reasoning ability, response speed, and Token consumption. Even if a model performs outstandingly in overall capability, it does not necessarily mean it can complete all work at the lowest cost.

Integrating more third-party models can expand the capability boundaries of Qianwen Office and allow users to choose more suitable models according to different tasks.

This means Alibaba needs to weigh the application needs of its self-developed models against the product competitiveness of Qianwen Office.

Over the past few months, changes in the external competitive environment have indeed made this multiple-choice question more urgent.

Tencent WorkBuddy had previously integrated products from multiple model vendors such as Hunyuan, GLM, MiniMax, Kimi, and DeepSeek.

ByteDance is also following quickly. On October 29 this year, Doubao Work took the lead in opening third-party models to team and enterprise subscription customers; on November 12, it further opened model selection to individual users and launched quota consumption multipliers and discount mechanisms.

Baidu's Kuku AI also already offers model choices such as Ernie, DeepSeek, GLM, Kimi, and MiniMax.

As competitors open up third-party models one after another, if Qianwen Office continues to insist mainly on self-developed models, it will need to rely on Qwen's performance in different office tasks to compete with the entire model lineup aggregated by competitors.

Even if Alibaba has strong self-developed model capabilities, it is difficult to guarantee that Qwen can balance effectiveness, speed, and cost in all tasks.

If users switch to competitors because they need to use other models, what Qianwen Office may lose is not just a single model invocation, but also subsequent office tasks and usage habits.

This reopening of third-party models also means that Qianwen Office has begun to place more emphasis on the platform's own competitiveness, rather than only on the invocation needs of self-developed models.

This change is also reflected in specific product design. According to the official introduction, the newly added Qwen4-Max is mainly aimed at multi-step complex reasoning tasks; models such as GLM-5.3 and DeepSeek V4 Pro can be used for analysis, judgment, and plan checking; while relatively low-cost models such as Qwen3.8-Flash, GLM-5.3-Flash, and DeepSeek V4.1 Flash are suitable for daily work such as copywriting and information extraction.

For enterprise customers, Qianwen Office also provides model management permissions. Administrators can configure the models available by department, user group, or member, matching different capabilities and costs to different positions.

The capability differences among different models are beginning to be translated into product choices for specific work scenarios.

However, with Alibaba joining, the multi-model competition in office Agents will also enter a new stage.

In the past, model capabilities and the scope of integration could still serve as differences between products; now, the model lineups each company can invoke are becoming increasingly similar, and simply integrating more models is no longer enough to establish a long-term competitive advantage.

The new differentiation will come from scheduling and execution capabilities above the models.

From the product information disclosed this time, Qianwen Office mainly emphasizes users' independent choice of models and has not set up an Auto mode; while Tencent WorkBuddy and ByteDance Doubao Work have further explored handing model selection to the platform through Auto mode.

There is an important difference between the two approaches.

Manual model selection requires users to judge task difficulty, model capabilities, and usage costs themselves; automatic model routing requires the platform to independently determine which model should be invoked based on task requirements and cost constraints.

The latter is obviously more difficult. A complex task often needs to go through multiple stages such as information retrieval, data analysis, content generation, tool invocation, and result checking. Different stages have different requirements for model capabilities, and how to control Token consumption while ensuring execution effectiveness has become an important factor affecting product experience and commercialization efficiency.

Beyond model routing, Agents also need capabilities such as task planning, context management, tool invocation, and error recovery. Together, these capabilities form the Agent Harness, which determines whether a task can be executed and delivered stably.

This also means that even if various office Agents use the same models, there may still be significant differences in final task completion rate, execution speed, and cost.

As model selection gradually becomes a standard configuration for office Agents, competition among vendors will shift further from which models are integrated to how models are scheduled and how tasks are completed.

Three months ago, Qianwen Office was still wavering between self-development and openness; now, competition in office Agents is making this multiple-choice question clearer.

Preserving the invocation volume of one's own models is certainly important, but preventing users' office tasks from flowing to competitors may be the more realistic consideration. Risk Disclosure and Disclaimer

Markets carry risks, and investment requires caution. This article does not constitute personal investment advice and does not take into account the special investment objectives, financial conditions, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their particular circumstances. Investment based on this is at the user's own risk.

Source

Wallstreetcn