At the Yunqi Conference in Hangzhou on Sept 22, major enterprises including Yihai Kerry's Golden Arowana and FAW Toyota announced adoption of Qwen Office; the same day, Qwen Office launched "Enterprise Context", a data management product for enterprise-AI collaboration β a clear step from general AI office tools toward deep enterprise customization.
π₯ Major Yunqi launch β Golden Arowana / FAW Toyota β "Enterprise Context" product π After 30M Qwen Office usersThe enterprise customer lineup around Qwen Office visibly upgraded at the Yunqi Conference: Golden Arowana β a household consumer brand in cooking oil and food, reaching hundreds of millions of families; FAW Toyota β a major joint-venture automaker with a large national dealer and supply-chain network. Combined with CIMC Enric (clean-energy equipment, business in 100+ countries), Qwen Office's enterprise references now span energy, manufacturing, consumer goods and automotive.
Adoption by industry leaders is a stronger signal than any keynote: enterprises of this size must pass four gates before buying an AI office product β data security, compliance, stability and integrability. Winning consumer and automotive giants at the same time means the product has cleared the enterprise bar, not just the "personal assistant" bar.
Enterprise adoption timeline (September 2026)
Earlier, companies including Changan Automobile, Huifu, Transfar Group and Laoxiangji had already onboarded. From manufacturing to consumer brands, from SOEs to private enterprises, adoption is shifting from individual lighthouse accounts to industry-wide rollout.
The biggest pain point in enterprise AI office is not model capability β it's that models don't "know" your company. Your policies, product docs, customer data and project conclusions live in different systems. Without that context, a general chat model gives generic answers; enterprises need answers grounded in their own data.
"Enterprise Context" is positioned as a data management product for enterprise-AI collaboration. It addresses three things:
| Layer | Problem with traditional approaches | Enterprise Context approach |
|---|---|---|
| Data access | Docs, sheets and system data are scattered; humans manually prepare content for AI | Manage enterprise data sources uniformly; AI reads within controlled boundaries |
| Context management | AI conversations are fragmented; no knowledge is retained across sessions or teams | Persist enterprise knowledge, business rules and past decisions as reusable context |
| Security boundaries | Employees paste sensitive data to AI with no governance | Enterprise-grade permissions and compliance; who sees what context is managed and auditable |
In short, "Enterprise Context" fills the middle layer that turns AI office from a personal productivity tool into an organizational system. Without it, AI writes emails and makes spreadsheets. With it, AI behaves like a colleague who actually knows the company β answering from real company data and executing tasks that match company workflows.
Why it appeals to large enterprises: A consumer company like Golden Arowana has knowledge spanning channels, distributors, marketing and quality control; a carmaker like FAW Toyota spans R&D, supply chain and after-sales. Making AI "know" this is a direct source of cost reduction and a step toward treating data as an asset. Enterprise Context turns "AI using company data" from an ungovernable gray zone into a governable enterprise capability.
Put the three September milestones together and the direction of the market is clear:
If your organization is evaluating AI office or enterprise AI, use this order:
For SMBs that want a fast validation, start with the free/team tier: put one real workflow into AI office (customer records, policy Q&A, weekly report summaries), then evaluate the enterprise tier. LLM capabilities are exposed through the Alibaba Cloud Model Studio stack, with up to 1M free tokens per model for new users β enough for a low-cost technical proof of concept.
Qwen Office is the product-layer packaging of Alibaba Cloud's LLM capabilities. If your team wants to build "Enterprise Context"-style features into your own systems β knowledge-base Q&A, RAG over enterprise data, internal agents β the foundation is Alibaba Cloud Model Studio:
Prices and promotions follow the latest Alibaba Cloud official pages; new-user benefits require real-name verification.
Related direction, but more systematic. Enterprise Context manages the full loop of how enterprise data becomes AI context β data access, knowledge persistence, permission boundaries and cross-session reuse. Think of it as the capability layer that makes AI "understand the company", not just a Q&A box over documents.
Yes. Qwen Office offers a free personal tier. SMBs can start on the team tier, validate with a real workflow (policy Q&A, customer records, weekly reports), then consider the enterprise tier for security, compliance, permission layering and usage governance at scale.
Adoption by large enterprises such as CIMC Enric (manufacturing), Golden Arowana (consumer) and FAW Toyota (automotive) is itself a validation of enterprise-grade security. For sensitive data scenarios, use the enterprise tier and rely on the official security whitepaper and contract terms.
Individuals and small teams can register and try it directly. For enterprise procurement, annual resources, GPU servers or Model Studio enterprise onboarding, use our Alibaba Cloud channel for the latest campaigns and channel pricing: SMB coupon campaign ο½ ECS & Model Studio.
If you're a decision-maker: add enterprise AI office to your shortlist and evaluate on data access, context management, security/compliance and foundation ecosystem. If you're a technical team: use the free token credits to run one knowledge-base Q&A or agent proof of concept β the smallest possible test of whether this path works in your scenario.
Qwen Office free tier Β· Up to 1M free tokens per model on Model Studio Β· Cloud servers from ~20% off for new customers
Explore Alibaba Cloud β Claim SMB Coupons