Traffic is the lifeblood of the digital age—why should it make way for Doubao?

Author｜Dingshan

Editor｜Xiaobai

Illustration｜AI-Generated

Produced by｜Qiangdiao Next

On September 16, the Nubia NaviX Ultra officially hit the market. With a starting price of 5,999 yuan, or 5,499 yuan after national subsidies, Nubia announced that its first-day sales surpassed 100 million yuan in just one second. Dubbed the 'second-generation Doubao Phone,' this consumer-grade device once again sparked expectations for 'phones that handle tasks for users.'
However, some of these expectations were dashed upon actually using the apps. Media outlets such as the National Business Daily conducted real-world tests, instructing the phone to open WeChat and post a moment. While the assistant could launch the app, it couldn't complete in-app actions. Services like shopping on JD.com, posting on Xiaohongshu, placing orders on Meituan, or shopping on Taobao were all restricted.

A joke began to circulate: calling it a 'ByteDance-exclusive machine.' The unfortunate truth? It was accurate.

From the 3,499-yuan first-generation prototype to the 5,999-yuan second-generation consumer model, the starting price increased by 2,500 yuan. Yet, the promise of cross-app functionality remained unfulfilled. Consumers who purchased the phone still had to manually perform their most frequently used services.

Doubao aims to operate the entire phone on behalf of users, but the platforms within the phone have their own agendas. Just because the model can recognize buttons doesn't mean ByteDance has the authority to press them on users' behalf. Even if phone manufacturers are willing to cooperate, it doesn't guarantee that WeChat, Taobao, or Meituan will follow suit.

The vision for the Doubao Phone is appealing: users state their needs, and AI navigates across apps to get things done. The reality, however, is stark: the steps it seeks to eliminate are precisely where platforms control users, allocate traffic, and generate revenue.

First Generation Hits a Wall in Four Days, Second Generation Starts Negotiating Rules

On December 1, 2025, the technical preview version of the Doubao Phone assistant was released, debuting on the limited-release Nubia M153 prototype priced at 3,499 yuan, which sold out on the same day. ZTE's stock price surged to the daily limit. Two days later, users reported abnormal WeChat logins. By the evening of December 3, Doubao announced the removal of WeChat operation capabilities from its phone assistant.

On December 5, Doubao further announced standardized adjustments, covering scenarios like incentive abuse, financial apps, and gaming. Just four days after its release, the scope of operations was voluntarily narrowed. The seamless path shown in demo videos encountered obstacles in the real internet, where each app could set its own barriers.

Nine months later, the second-generation Doubao Phone responded by codifying app willingness into rules. On September 14, Doubao introduced the Screen Automation Operation Statement Protocol (SAEP), allowing third-party apps to declare whether they accept GUI agent operations. The rules unfold in two phases: during the 30-day public comment period, apps are not operated by default and must opt-in to be accessible; after the period, apps that haven't explicitly declined via email or protocol will have operations enabled. Apps that explicitly refuse will not be automatically operated.

These rules provide a clear negotiation window for all parties, revealing how Doubao's capabilities are determined: ByteDance develops the functions, but apps decide whether to enable them.

The initial restrictions include default arrangements during the public comment period, but long-term accessibility still hinges on app platforms' attitudes.

For users, this means the most attractive selling point of the Doubao Phone must await cooperation progress to materialize.

Apps Want Orders, Not to Surrender Entry Points

Restrictions between phone manufacturers and apps are not absolute. On July 13, StepOn released the STEPX terminal brand and STEPX Neo phone solution, announcing initial ecosystem partners including Ctrip, Alipay, Didi, Meituan, JD.com, and WPS.

In June of this year, WeChat also began collaborating with Huawei, Honor, Xiaomi, OPPO, and Vivo on A2A (App-to-App) capabilities. Media reports indicate that Honor's YOYO can now initiate WeChat audio/video calls and send messages to specified contacts.
These cases demonstrate that app platforms accept AI participation in services, but only after negotiating what AI can do and how it's invoked.

App platforms welcome an additional channel that brings orders but are reluctant to embrace a superior that allocates orders for them.

When a user says, 'Help me buy headphones under 300 yuan,' the AI assistant can either enter a designated platform to use its products and sorting or search across platforms to compare prices, brands, and after-sales service before deciding where to place the order. The clicks saved shift the most valuable decision-making power before transactions.

Advertising relies precisely on this decision-making power. Tencent's marketing service revenue grew 22% year-on-year in Q2 2026, partly driven by AI enhancing ad efficiency within the WeChat ecosystem. Internet companies are using AI to sell their traffic more effectively, naturally calculating the cost of external assistants taking over user choices.

ByteDance is a uniquely influential negotiator. With e-commerce and local services already established through Douyin, Doubao gaining insight into shopping needs, budgets, and preferences creates another avenue to influence transaction flows. For other platforms, granting operation permissions simultaneously strengthens a competitor's hand—an obvious contradiction.

Doubao aspires to be users' chief assistant, but internet platforms prefer it as just another service channel. The core conflict lies in who decides the next step after a user states a need.

This negotiation won't end automatically with increased model parameters. ByteDance must present transaction growth, allocation rules, and cooperation terms that make financial sense for platforms.

382 Million MAUs Can't Buy System Permissions

Why does ByteDance push ahead with a second-generation phone despite knowing the resistance? QuestMobile data shows Doubao had 382 million monthly active users (MAUs) in June 2026, more than double Qianwen's 167 million. Yet no matter how many users Doubao has, it remains constrained by system permissions and app collaboration conditions on phones.

A chat app can wait for users to open it before answering questions. A task assistant must be summoned instantly, understand task progress, operate different services sequentially, and take over when errors occur. Whoever controls the system decides how these capabilities integrate into daily use.

Partnering with ZTE on this Doubao Phone provides ByteDance a testing ground: where the assistant is summoned, how tasks execute, and how preferences are remembered can all be deeply involved in design. ByteDance aims to secure definition rights for next-gen phone interactions. Otherwise, even if Doubao models enter more phones, they might merely serve as suppliers behind manufacturers' proprietary assistants.

The problem? Phone manufacturers are also vying for this position.

AI assistants that learn user habits, connect devices, and allocate services are tools for manufacturers to differentiate products. Procuring Doubao models can enhance their assistants, but ceding the default assistant role to Doubao lets ByteDance directly control users.

App platforms guard service entry points; phone manufacturers guard system entry points. Doubao must pass through both to directly control user choices.

Nubia offers a window, but its terminal scale remains tiny. Media reports indicate Doubao AI phones have about 200,000 units in stock, with initial batches under 100,000. IDC data shows the top six Chinese smartphone vendors accounted for ~96% of Q2 shipments: Huawei 22.6%, Apple 18.1%, OPPO and Vivo 16.0% each, Xiaomi 12.4%, and Honor 11.3%. Even counting all ZTE shipments, its market share is just 0.3%.

This poses a dilemma for ByteDance: only with complete services can users have stronger reasons to switch phones. Only with sufficient users will service providers invest in adaptation. While Doubao has 382 million MAUs as an app, that doesn't directly translate to phone users. WeChat's 1.4 billion users don't guarantee success for WeChat-customized phones either.

ByteDance can subsidize hardware to acquire users, but WeChat's social graph, Meituan's supply network, and Taobao's merchants won't arrive with a single hardware subsidy.

Users Aren't Obligated to Wait for Negotiations

Even if all apps cooperate, consumers might not immediately gain a reliable task assistant. Another challenge for Doubao is how much it can truly save users from doing.

AndroidDaily research released in May 2026 tested 350 daily tasks across 94 real Android apps, finding the strongest model succeeded in 62% of cases. This benchmark highlights the difficulty of sequential operations: recognizing a button differs from completing a task.

No tickets available, ambiguous product specifications, or unusable coupons all require the assistant to rejudge. Users don't care how many steps it completes; they care about whether the final purchase or booking is correct and more convenient than doing it themselves.

In real-world Doubao Phone tests, the second generation still required manual handover during payment and identity verification. For the product, smooth handovers and error corrections matter as much as app accessibility in determining experience.

As cooperation expands, another constraint emerges: the assistant's choices are limited by the cooperation scope.

The same request—'Help me book a hotel'—can mean comparing prices and cancellation policies across platforms or merely selecting from within cooperative channels, both packaged as one-sentence bookings. Users tap fewer screens but might not get better options.

If Doubao can only operate within negotiated cooperation boundaries, it must prove this scope fulfills its promise as a 'universal assistant.'

From the first to second generation, Doubao Phone has showcased the intensity of the AI entry-point battle: system permissions must be secured, platform interests coordinated, and task execution made reliable. ByteDance can view these investments as costs for securing the future, but consumers pay for today's product.

A phone priced starting at 5,999 yuan should convince users with tasks it can already complete. The 'one-second billion' in first-day sales generated buzz, but whether WeChat, shopping, and food delivery work reliably long-term will determine how many users stay.

Doubao Phone's vision is beautiful—letting users navigate the entire internet with one sentence. But first, it must resolve why internet companies should let it cross their domains.

Note: Data in this article comes from public sources and does not constitute investment advice.

- END -