Alibaba released the proprietary Qwen3.6-Plus and accelerated Qwen's pivot toward monetization
Alibaba released the proprietary Qwen3.6-Plus and, in just three days, unveiled three proprietary models. For a company that spent a long time building…
AI-processed from 3DNews AI; edited by Hamidun News
Alibaba presented its closed model Qwen3.6-Plus and signaled at the same time that the strategy around the Qwen family is changing. In three days, the company released three closed AI models at once — a notable shift for a player that was previously primarily associated with open releases.
New Course for Qwen
Until now, Alibaba has actively strengthened the Qwen brand through open models that quickly spread across GitHub, cloud platforms, and corporate pilots. This strategy helped attract developers, build community, and establish Qwen among the most prominent AI lineups from China. But the launch of Qwen3.
6-Plus in closed format shows that the company no longer wants to limit itself to the role of technology supplier for the ecosystem and is increasingly building a commercial product vertical. The meaning of this shift lies not just in the single model, but in the pace of releases. When three closed products come out in three days in a row, this is no longer a test or a single experiment.
It's a signal that Alibaba is dividing the lineup into two circuits: open — for distribution, recognition, and external integrations, and closed — for flagship scenarios where control, differentiation, and revenue matter.
Why Close Models
A closed model gives a company a completely different set of levers than an open release. If weights and architecture are not published, Alibaba can update the product faster, hide some optimizations, manage access more precisely, and sell premium capabilities as a service. For large business this is also a clear format: not a set of files and instructions, but a managed API or cloud product with predictable quality and SLA.
- Premium access to better models is easier to monetize
- Architectural improvements and training recipes stay within the company
- Corporate clients find it easier to buy a service than assemble the stack themselves
- A closed circuit simplifies security control, rate limits, and updates
- Flagship models can be more strictly differentiated by tier and use case
Importantly, such a move does not necessarily mean abandoning open source. Rather, Alibaba is showing a multi-level strategy: open Qwen models continue to work as a technology showcase and distribution channel, while closed versions become the top layer where the company wants to capture the main margin. For the market, this is already a familiar scheme: first ecosystem and reach, then paid access to the strongest or most convenient versions.
What the Market Will See
For developers, this shift means that some of the most interesting capabilities within Qwen may move to a publicly inaccessible layer. If before the brand's value was built on the fact that models could be downloaded, fine-tuned, and embedded in your own stack, now the emphasis shifts to a model as a service. This doesn't kill interest in the open family, but it changes expectations: best qualities may increasingly remain on the Alibaba cloud side.
For competitors, the news is also revealing. The AI market quickly moved to a model where open source helps win attention, and closed products — money. If Alibaba cements this scheme, Qwen will ultimately turn from just a research brand into a full-fledged commercial platform.
And the release of three closed models in three days looks like a demonstration that the company is ready to accelerate this transition rather than drag it out over quarters.
What This Means
Alibaba is not abandoning the Qwen legacy as an open ecosystem, but is increasingly moving the upper segment of the lineup into a paid and controlled format. For the market, this is another sign: in generative AI, the competition is no longer just about model quality, but about who can best turn them into revenue.
Want to stop reading about AI and start using it?
AI News is a curated feed of AI/tech news. Hamidun Academy teaches you to use AI systematically in your work.