On July 15, 2026, Chinese regulators published a list of seven on-device generative AI services that had completed the required filing. Apple Intelligence was among them. The Chinese build differs from the US version: Alibaba's Qwen handles the language side, while Baidu covers search and visual recognition. After nearly two years of standstill, Apple's AI features have finally cleared the regulatory entrance to the mainland market.

Nearly Two Years of Waiting

Apple first showed off its AI features in 2024. Mainland China stayed off the list of supported regions ever since, leaving a gap of roughly 22 months. The reason was straightforward: anyone offering generative AI services to the public in China must file with the authorities first. Bringing an overseas model in unchanged is effectively impossible, so pairing up with a local company is the price of entry.

The filing published this time lists Apple's Chinese subsidiary as the registrant, with iPhone named as the applicable scenario. It is worth stressing, though, that the notice only confirms the paperwork is done. No date has been given for when the features actually become available.

Alibaba on Language, Baidu on Search

The interesting part of the Chinese setup is that Apple did not lean on a single partner.

Alibaba's Qwen takes the language side. It serves users of iOS, iPadOS, macOS, and visionOS in China with text and image understanding and generation. From the user's point of view, Qwen's capabilities are reachable from within Apple's own interface, with no need to open a separate app. On the news, Alibaba's US-listed shares rose by close to 4 percent at one point.

Baidu takes the other half: image-based search and the part where Siri reaches out to the web. In the US version, Google's Gemini and OpenAI fill those roles. In China the work is split between two companies, which can also be read as Apple making sure no single firm ends up holding its China AI stack.

iPhone 15 Pro and Later, Starting with iOS 26.4

Running the Chinese version of Apple Intelligence comes with conditions. According to reports, it requires iOS 26.4 or later, an iPhone 15 Pro or newer, and an Apple ID registered in the China region.

The device requirement is limited to the Pro line and later because most of the processing is designed to stay on the device. That fits with the fact that this filing was made under the category of on-device generative AI. Running inference locally rather than shipping data to the cloud was presumably an easier proposition for regulators to accept.

Seven Vendors Cleared at Once

It is easy to miss, but this was not only about Apple. The same notice covers Huawei's Xiaoyi, OPPO's AndesGPT, vivo's BlueLM, Xiaomi's HyperAI, SAMSUNG's Galaxy AI, and nubia's Doubao model. In other words, on-device AI from the major handset makers stepped onto the regulatory stage together.

Filings in China have so far centered on large models running in the cloud. This is reported to be the first notice aimed specifically at AI running inside phones, which marks a shift for on-device AI from an experimental feature to one embedded in the regulatory framework. For Apple, it means a long-missing piece in one of the world's largest markets is finally within reach.

Summary

Apple Intelligence has cleared the filing process in China and is set to ship with Alibaba's Qwen handling language processing and Baidu handling search and visual recognition. Running it reportedly requires iOS 26.4 or later, an iPhone 15 Pro or newer, and a China-region Apple ID. No launch date has been announced yet, so how well it actually works remains to be seen.