News at a Glance
Google says the AI-calling feature will first be available to Pixel 11 owners in the U.S. who pay for a Gemini subscription.
What Happened
In an official announcement on September 24, 2026, Google confirmed that the AI-calling feature will initially be available to Pixel 11 users in the U.S. who hold a Gemini subscription. The feature is not enabled by default but is offered as an add-on benefit, requiring users to pay a monthly or annual fee. Pixel 11, launched in fall 2026, features the third-generation Tensor processor, with local AI compute power up 40% from the previous generation, providing hardware support for real-time speech recognition, semantic understanding, and generative responses during calls. Google did not disclose the full technical architecture, but industry consensus suggests a hybrid app
Background
This launch occurs amid intensifying competition in AI-powered smartphones. Since 2025, Samsung, Apple, and Huawei have accelerated integration of generative AI into system-level applications, but most focus on non-real-time scenarios like text generation, image editing, or voice assistants. Real-time AI intervention during calls has rarely been commercialized at scale due to privacy, latency, and misrecognition risks. Google has long bet on AI differentiation for Pixel, and its Gemini subscription service, launched in 2024, has built a user base, though
Deep Dive
Liu Gong believes the technical challenge of AI calling lies not in model capability but in balancing real-time performance and privacy. Call scenarios demand millisecond-level response; any delay over two seconds disrupts conversational flow. Google's hybrid approach—on-device processing for transcription and basic intent recognition, cloud processing for complex semantic generation—leverages Tensor G3's local compute while preserving Gemini's reasoning power. Liu Gong predicts actual latency on Pixel 11 will stay under 1.5 seconds, but in poor network conditions, the system may degrade to pure on-device mode, reducing response quality. Moreover, call content involves highly sensitive personal information; Google must find a regulatory-acceptable balance among user authorization, data encryption, and local storage, otherwise the feature could become a privacy controversy trigger. Liu Gong points out that Google's core intent is to enhance Gemini's perceived value. The current AI subscription market is highly homogeneous, making it hard for users to distinguish paid from free tiers. By making AI calling subscription-exclusive, Google effectively places a paywall in the most frequently used scenario, creating stronger payment incentives than merely increasing model parameters or generation limits. Liu Gong estimates that if Pixel 11 sells 5 million units in the U.S. and Gemini conversion rises from 15% to 25%, this feature alone could generate approximately $125 million in annual incremental revenue. However, the risk is equally clear: if users perceive call AI as a basic function rather than a premium add-on, negative backlash could damage Pixel's brand reputation. Liu Gong argues this move will force competitors to reassess AI monetization strategies. Samsung's Galaxy AI is partially free and partially paid, while Apple Intelligence is entirely free; Google's subscription model could become an industry watershed. If the market accepts this model, other vendors may follow, bundling high-frequency features like AI calling and real-time translation into paid plans. If users resist, it may reinforce the consensus that AI should be free. Liu Gong predicts that within 12 months, at least one major Android vendor will launch a similar subscription-bound feature, while Apple will likely maintain a free strategy to consolidate ecosystem loyalty.
Data Points
- Compared to Pixel 10, Pixel 11's Tensor G3 processor delivers 40% more AI compute power, with on-device model parameters rising from 7 billion to 12 billion, improving real-time speech recognition accuracy from 92% to 96%. Against competitors, Samsung Galaxy S26's AI calling only supports transcription and summaries, requiring manual activation, while Google's AI calling offers proactive suggested replies with a shallower interaction layer. Market-wise, as of Q2 2026, Gemini subscription users numbered approximately 80 million, with U.S. users accounting
Impact & Outlook
For the industry, in the short term (3 months), Google's subscription bundling will spark observation and debate within the Android camp. Samsung and Xiaomi may accelerate evaluations of monetizing their own AI features but are unlikely to follow immediately, to avoid user backlash during flagship launch cycles. In the medium term (1 year), if Google publishes user retention and conversion data for the feature and performance exceeds expectations, a wave of subscription-based AI features may emerge, potentially marginalizing free AI offerings. For ordina
Angles to Explore
Points to consider
- Privacy boundaries of AI calling: How will data processing be split between on-device and cloud, and how will regulators intervene?
- Commercial sustainability of subscription-based AI: Can user willingness to pay support Google's long-term AI investment?
- Is Pixel 11's hardware upgrade sufficient for real-time AI calling, or does it rely excessively on cloud processing?
Source & Original
This item comes from TechCrunch AI (published 2026-09-24 16:00:00). Wenfeng Radar provides a Chinese/English summary and commentary on overseas AI developments; the original content is owned by its author.
Daily aggregation of overseas AI developments with in-depth analysis. Bookmark us to never miss an important signal; back to homepage.
