When Apple unveils its significantly enhanced version of Siri in 2026, the artificial intelligence at its core will reportedly have a major dependency on its rival, Google. According to a new report from Bloomberg, Apple is finalizing plans for a strategic partnership to use a custom version of Google’s Gemini model to power key functions of the modernized voice assistant, paying a hefty sum for the privilege.
The overhauled Siri will operate using a hybrid approach, leveraging both Apple’s own smaller, homegrown models and a specialized version of Gemini. The Google model will be critical for handling Siri’s “summarizer and planner functions,” which involve synthesizing information and determining the best way to execute complex user requests. Since using apps on a user’s behalf is a core capability of the updated assistant, Gemini is set to play a foundational role in enabling these complex tasks.
To facilitate this partnership while maintaining its strict privacy standards, Apple will run the Gemini model exclusively on its own secure infrastructure: the Private Cloud Compute servers. This arrangement addresses Apple’s concerns about data security and control, which were reportedly central to the negotiations.
The financial terms of the deal are significant. Apple is expected to pay Google an estimated $1 billion annually for access to this custom Gemini technology. While this figure is substantial, it is noted to be only a fraction of the amount Google reportedly pays Apple to maintain Google Search as the default search engine across Apple devices.
Despite the high-profile nature of the partnership, it is expected to be a collaboration that Apple will not actively advertise. The iPhone maker is simultaneously working toward developing its own proprietary solution, viewing the Gemini integration as a necessary, yet temporary, step to catch up in the rapidly accelerating AI race.
Apple’s ultimate goal is to replace the Google-backed model entirely. Bloomberg reports that the company hopes to have a custom, one trillion parameter cloud-based model ready for consumer applications as early as next year, which would allow the company to achieve full independence in its core AI capabilities.
The pursuit of an external partner follows a period of reported challenges for Apple’s internal AI development. While Siri can currently integrate ChatGPT for specific requests as part of the broader Apple Intelligence suite, the need for a more robust partner became evident after Apple delayed Siri updates in 2025.
Explorations for this partnership were first rumored in August, when Google was tasked with creating a Gemini version that could operate securely on Apple’s servers. Before settling on Google, Apple was also rumored to have explored potential deals with other major AI developers, including Anthropic and OpenAI.