What Did Alibaba Really Buy After Qwen Was Integrated Into iPhone?

By emphasizing Next

On October 31, Apple officially announced Apple Intelligence for China, with testing opening on November 4 alongside the developer Beta and a rollout in December with iOS 27.3, bringing Qwen into Siri, Writing Tools, and Visual Intelligence. Most of the first batch of features are ones that have been rolled out overseas since 2024; the next-generation Siri AI is not among them.

Around this news, most of the market's interpretation was: with more than 200 million existing devices, system-default invocation, and zero-cost reach, Alibaba Cloud's Token revenue would expand accordingly. When news of the filing emerged in July, Alibaba's US shares briefly rose more than 7%, and brokerages also raised their revenue forecasts for Alibaba Cloud.

But the product arrangements announced in October are already clear: Qwen is an extension that users choose to enable on their own. How many existing iPhones can be covered, how many of those users will use it, and how these invocations will be settled are all unknowns.

At list prices, this business is worth only a few hundred million to a few billion yuan a year

First, look at the brokerages' numbers. Because of this partnership, SDIC Securities International raised its FY27 revenue forecast for Alibaba Cloud by 4%, corresponding to a 47% growth rate in cloud intelligence revenue. Alibaba Cloud Intelligence's FY26 revenue was 158.1 billion yuan; based on a 47% growth rate and a 4% upward revision, the implied increment is roughly 9 billion yuan. But without the breakdown from the original research report, this 8.9 billion yuan cannot all be attributed to the Apple partnership.

Next, look at devices. The first batch supports only the iPhone 15 Pro and later models. IDC data shows that Apple shipped 46.2 million units in China in 2025, up 4% year on year, implying about 44 million units in 2024. The two years together total about 90 million units; adding the installed base of the 15 Pro and new phones in 2026, the eligible models at the December launch are around 100 million units. To observe the order of magnitude, this article temporarily uses 100 million eligible in-use devices for scenario calculations.

For a single invocation, assuming 1,500 input Tokens and 400 output Tokens, the cost of image understanding has already been included. The list price for the Qwen Plus tier is about 2 yuan per million input Tokens and 12 yuan per million output Tokens, equivalent to about 0.8 fen per invocation. The flagship Max tier is 12 yuan for input and 36 yuan for output, equivalent to about 3 fen per invocation. For tasks like summarization, rewriting, and image recognition on Apple devices, the Plus tier is more reasonable, with the Max tier as an upper-bound reference.

Annual inference amounts under three scenarios are as follows:

Even under the aggressive scenario at Max-tier list prices, the amount is only about 3.5 billion yuan, still 2.6 times short of the 9 billion yuan implied by the brokerages. To match that number, one would have to simultaneously assume 30% activation, 10 uses per day on average, all using the flagship model, and Token consumption magnified by another 2.6 times. By comparison, Alibaba's AI cloud and computing power services revenue in the June quarter was 48.4 billion yuan, with capital expenditure of 67.7 billion yuan in the single quarter. List prices are retail prices; Alibaba's own internal costs are lower.

Whether inference at this layer is recorded as revenue or cost, it falls within the rounding error of Alibaba's financial reports. The earlier claims of "zero-cost reach" and "volume expansion" need to be reframed: if this partnership is not worth much money, why is Alibaba still doing it?

The direction of money is most likely reversed

Look at how Apple's past two AI deals worked. According to Bloomberg, Apple pays Google about $1 billion a year to use a custom Gemini with 1.2 trillion parameters to power the new Siri, with the model running on Apple's own private cloud. In January 2026, the two sides formally confirmed a multi-year partnership: the next generation of Apple foundation models will be based on Gemini models and cloud technology, continuing to run on Apple devices and its private cloud computing system, but financial terms were not disclosed.

Money flowing in the opposite direction is much larger. Court documents show that Google paid Apple $20 billion in 2022 for default search in Safari, and a Google witness also disclosed that Google shares 36% of search advertising revenue with Apple (Apple has a similar arrangement with Baidu in China).

When ChatGPT was integrated into Siri in 2024, Bloomberg reported that Apple did not pay OpenAI, believing that putting the latter on hundreds of millions of devices was worth no less than cash, and it planned to charge commissions in the future from AI partners that monetize on Apple's platform.

Apple's price list is very clear: for a "brain" it cannot build itself, it pays. For an "exit" that brings traffic, it charges. Qwen's current position is the exit: the first batch of features is powered by Apple's own models, and Qwen is an extension that users can optionally enable, and users also have to log in to a Qwen account.

In this position, money is more likely to flow from Alibaba to Apple. Among the three revenue paths inferred by brokerages, both fixed licensing fees and per-Token payments require Apple to pay, which runs counter to its practice. Only the e-commerce commission path aligns with its practice, and the party receiving the commission is precisely Apple.

Apple also has no incentive now to pay on behalf of partners. Its gross margin guidance for the fourth fiscal quarter is 47% to 48%, below market expectations, and Cook said it is evaluating all options to ease cost and supply pressures. The background is the continued surge in prices of NAND and LPDDR (flash memory and memory chips). Making partners bear cloud inference is Apple's most convenient choice.

Even if Qwen someday takes the "brain" position of Siri in China, it will not be a big sum of money. Greater China accounts for about 17% of Apple's total revenue ($18.8 billion out of $109.4 billion). Scaling Google's $1 billion price by this ratio gives about 1.2 billion yuan a year.

What Alibaba is buying is the qualification to have both the "brain" and the "hands" present

If the money is not worth much, why is Alibaba still doing it? The answer lies in the Doubao phone.

On October 14, the rejection list disclosed by the Doubao phone assistant comprised 155 apps in total, covering WeChat, Taobao, Alipay, Meituan and others, among which 66 were Tencent-affiliated and 23 Alibaba-affiliated, with Taobao, Alipay, Amap, Hema and even Qwen itself all on the list. Some analysts believe that high-barrier domains such as social networking and payments are more likely to refuse to hand over control to a system-level Agent, and instead build app-level Agents within their own apps. Doubao therefore had to switch to partnership interfaces, and has so far integrated willing services such as Caocao Mobility, and also launched its own "Doubao Shopping."

On phones in China, the scarce resource for Agents is not the model, but the qualification to be allowed through by high-frequency apps. Models can be swapped; the app's door has to be opened by the app itself.

Alibaba holds a position in Apple's candidate list that no one else has: it simultaneously controls Taobao, Alipay, Amap and Ele.me—an entire set of transaction, payment, map and local-life apps. Tencent holds WeChat, but in currently public information Apple's partner is only Alibaba (and reportedly Baidu). For the China-market Siri AI to be truly useful, it must let people complete booking, buying, paying and navigation in one sentence, and most of these actions fall within Alibaba-affiliated apps.

If Qwen becomes the brain behind Siri, the way these apps open up to it is through partnership interfaces, without needing to force its way in via screen-reading and simulated clicks like Doubao. However, this is the inference of this article; no party in public information has mentioned any interface arrangement between Siri AI and Taobao or Alipay.

English text is being prepared.

This card cannot be played today

Alibaba holds the card, but Apple has no reason to rush to make an offer. Counterpoint data shows that in the quarter ending in June, Apple's iPhone sales in Greater China grew 23%, with share rising from 13.9% to 18.1%. This growth occurred during the two years when the China-market iPhone had no Apple Intelligence, though Greater China revenue was below analysts' expectations. Apple does not rely on this feature to sell iPhones; China-market AI is a defense line, not a growth engine. The less anxious Apple is, the less bargaining power Qwen has, which explains why Qwen only received an optional extension slot.

On the user side, there is the tuition ChatGPT has already paid. In a court filing on September 22, OpenAI stated that after ChatGPT was integrated into Siri, it got off to a slow start in January 2025 and was still described by March 2026 as continuing to underperform expectations, with the specific figures redacted. Reports point out that one reason was that the integration was off by default. Qwen's integration approach is almost the same template.

Qwen likewise needs to cross the distance from "can be used" to "is continuously used." System coverage scale is only the starting point; whether users turn it on, how many times they invoke it each month, and whether they come back determine the actual value of this entry point.

Now look at the denominator. With a base of 100 million eligible devices, and assuming a 10% to 30% activation rate, that is 10 million to 30 million people. In September, Qwen's monthly active users were 158 million, down from 167 million in June; over the same period, Doubao rose from 382 million to 419 million, a net increase of about 37 million in three months. Even at its best, the iPhone channel would only equal one quarter of Doubao's net increase and cannot change the Qwen-versus-Doubao landscape.

Structural erosion is still to come. Apple's approach prioritizes on-device processing, keeping basic tasks on the device and having partners handle only the parts that cannot be handled locally. With each advance in on-device models, partners are invoked fewer times.

The biggest variable is who owns Siri's brain. Apple said back in 2024 that it hoped users could choose among different models and named Gemini; two years later, the foundation of the new Siri was handed to Google. As for who will be the brain of the China-market version, industry observers expect the answer to come only in 2027 or even later.

Qwen now stands exactly where ChatGPT stood in 2024. China has fewer eligible compliant partners than the United States; Alibaba has filings, cloud computing power and an app cluster, so its opportunity is greater than ChatGPT's was then, but no one has signed yet.

On November 3, Apple will hold its earnings call, and Ternus's and management's remarks on China-market AI will be worth watching. After the December rollout, the more valuable data will be the actual activity, invocations and retention of the Qwen system extension. If a transaction interface is subsequently announced, it will also be necessary to look at completion rates and conversion, not just the integration list.