What happened
Qwen Office and Zhipu AI have initiated offline promotion campaigns in Hangzhou to distribute AI token credits directly to local entrepreneurs and developers. Qwen Office, in cooperation with the Hangzhou Talent Management Service Center, is distributing 1,000 token cards to 35 municipal-level college student entrepreneurship parks starting September 20. Each card contains 10,000 points, valued at approximately 500 yuan, which employees can redeem for Qwen Office usage. Simultaneously, Zhipu AI launched a specific 'GLM Coding Plan' event targeting developers in the same region. These initiatives represent a strategic pivot from purely digital marketing to grassroots, offline user acquisition, treating AI tokens as a standard production cost for startups similar to rent or utilities.
According to 36Kr, Qwen Office and the Hangzhou Talent Management Service Center began distributing AI token cards on September 20 to employees in 35 municipal-level college student entrepreneurship parks in Hangzhou. The first batch consists of 1,000 cards, each containing 10,000 points with a market value of approximately 500 yuan. Employees can obtain one card per person by presenting their work ID badges at the park management department. The process is designed to be simple, requiring no complex project applications or extensive documentation; users simply scratch off a redemption code to activate the points in the Qwen Office platform.
Concurrently, Zhipu AI launched an offline 'GLM Coding Plan' event specifically tailored for Hangzhou, targeting developers and users of AI coding tools. While these are distinct events, 36Kr notes they share a similar strategic logic: delivering tokens directly to developers and entrepreneurs to encourage initial product usage. This approach contrasts with traditional entrepreneurship subsidies, which typically focus on rent exemptions, free workstations, or utility bill reductions. The article argues that for modern AI startups, tokens have become a daily production cost, necessary for writing code, generating documents, running agents, and analyzing data.
The article highlights that Zhipu AI has also opened an official flagship store on Tmall, where it sells token and coding plan packages. Reports cited by 36Kr indicate that within 48 hours of opening, Zhipu AI's brand search volume increased by 50 times month-on-month, contributing to a 160% month-on-month increase in transaction volume for AI token and subscription products on Taobao and Tmall. This e-commerce presence complements the offline park distribution, creating a multi-channel strategy for user acquisition.
The source emphasizes that these moves are part of a broader trend where large model vendors are behaving like mature internet companies, focusing on user acquisition, retention, and conversion rather than just technological superiority. The article draws a parallel to the early cloud computing era, when Alibaba Cloud, Tencent Cloud, and AWS competed for startups by offering free server credits and usage quotas. The goal is to reduce the cost of switching vendors by embedding their services into the startup's workflow from the beginning.
Why it matters
This development signals that the AI industry is transitioning from a phase focused solely on model capability and benchmark scores to one driven by commercialization and user retention. By distributing tokens directly to early-stage companies, vendors are aiming to establish long-term usage habits and lock in customers before competitors can. This mirrors the early cloud computing market, where free server credits were used to capture startup workloads. The involvement of government entities in distributing these credits further legitimizes AI tokens as a core component of business infrastructure, potentially influencing future entrepreneurship support policies to include AI resource subsidies rather than just physical space or rent reductions.
The shift from digital-only marketing to offline grassroots promotion indicates that AI vendors recognize the importance of establishing deep usage habits among high-value users, such as developers and small business teams. These users consume significant amounts of tokens daily, making them critical for long-term revenue. By providing free or subsidized tokens, vendors aim to capture this segment before competitors can, effectively locking in customers through workflow integration.
The involvement of government entities in distributing AI tokens suggests a maturation of the AI industry's role in the economy. As tokens become a standard line item in startup budgets, similar to rent or electricity, government support systems may evolve to include AI resource subsidies. This could influence future entrepreneurship policies, potentially making AI access a key factor in determining the attractiveness of tech hubs and innovation parks.
This commercialization phase marks a transition from a 'technological war' focused on model capabilities and benchmarks to a 'commercial war' focused on market share and user retention. The success of these offline campaigns will likely determine which vendors can sustainably grow their user bases and convert free users into paying customers, shaping the competitive landscape of the AI industry in the coming years.
What to watch next
Monitor whether other major AI vendors in China or globally adopt similar offline distribution strategies in key tech hubs. Watch for changes in government entrepreneurship support policies to see if token subsidies become a standard benefit alongside rent exemptions. Additionally, track the long-term retention rates of users acquired through these campaigns to determine if this grassroots approach effectively converts free users into paying subscribers, and observe if e-commerce platforms like Tmall continue to expand their AI token subscription offerings.
Observe whether other AI vendors, both in China and internationally, replicate this offline distribution model in other major tech hubs. If successful, this could lead to a widespread trend of AI vendors partnering with local governments and entrepreneurship parks to distribute credits, potentially creating a new category of AI-related government subsidies.
Track the long-term impact of these token distributions on user retention and conversion rates. If users acquired through these campaigns remain active and convert to paid plans, it will validate the effectiveness of grassroots promotion as a key strategy for AI commercialization. Conversely, if retention is low, it may indicate that free tokens alone are insufficient to drive long-term engagement.
Monitor the expansion of AI token sales on e-commerce platforms like Tmall and Taobao. The reported surge in transaction volume suggests a growing consumer and business demand for packaged AI services. Future developments may include more competitive pricing, bundled offerings, and integration with other digital services, further normalizing AI tokens as a standard digital commodity.