Apple is preparing to launch a completely redesigned version of Siri around September 2026, likely along with iOS 27. This new Siri is expected to be a full rebuild rather than a normal update, turning it from a basic voice assistant into a modern AI system. But a key part of this upgrade is that Apple will not build everything alone – it is expected to use Google’s Gemini AI models for intelligence and Nvidia’s powerful AI chips to run parts of the system in data centers, reports The Information.
The most significant change in this project is the tech titan’s move toward using Google’s Gemini AI models to power Siri’s thinking and language abilities. The latest report suggests that the iPhone maker is paying around $1 billion per year for access to a customized version of Gemini, which is a very large model with about 1.2 trillion parameters. This is much more advanced than Apple’s in-house cloud AI models, which are estimated to be around 150 billion parameters. The move clearly allows Apple to quickly match the level of advanced AI systems already used by leading competitors without fully building everything from scratch.
The new Siri will work on a hybrid system. Simple tasks like setting alarms, checking basic information, and controlling device settings will run directly on Apple devices using on-device AI. But more complex tasks – like summarizing long documents, handling deep questions, or performing actions across multiple apps – will be processed in the cloud using Google’s Gemini models. Apple is also using a technique called model distillation, which helps it learn from Gemini’s responses so its own smaller models can improve over time.
Importantly, for cloud processing, Siri will rely on Google Cloud infrastructure combined with Nvidia’s AI hardware, particularly the newer Blackwell B200 GPUs. These chips are designed for large-scale AI workloads and are commonly used in data centers running advanced AI models. Apple is also expected to use Nvidia’s confidential computing features, which keep data encrypted even while it is being processed. This helps Apple maintain strong privacy protection even when user requests are handled outside the device.
The upgraded Siri is expected to become much more aware of personal context. It will be able to use information from apps like Mail, Messages, Calendar, Photos, and Notes to answer questions more intelligently. It will also support deeper app control, allowing users to complete multi-step actions like writing emails, editing documents, or managing tasks across different apps with a single command. Another major improvement is that Siri will become more conversational. Instead of responding only to one command at a time, it will maintain context across multiple messages, allowing follow-up questions without repeating information. This makes the experience closer to modern AI chat systems like ChatGPT or Google Gemini.
The Tech Portal is published by Blue Box Media Private Limited. Our investors have no influence over our reporting. Read our full Ownership and Funding Disclosure →
Ashutosh is a Senior Writer at The Tech Portal, largely reporting on new tech, and intersection of technology and business. Ashutosh’s career spans across nearly a decade of technology writing across multiple platforms and languages.
