Apple is poised to make a major shift in its Siri virtual assistant, with reports emerging that the company will be using a custom version of Google's Gemini AI model to power its updated voice assistant. Set to debut in 2026, the new Siri is expected to rely on both Apple-developed models and the Google-backed Gemini running on the company's Private Cloud Compute servers.
According to Bloomberg, the Gemini model will play a critical role in handling Siri's summarizer and planner functions, which enable the virtual assistant to synthesize information and execute complex tasks. Given Siri's core function of using apps on behalf of users, the partnership with Google appears strategic.
Sources indicate that Apple is willing to pay $1 billion annually for access to Gemini technology, a significant chunk of change compared to the rumored $10 billion payment made by Google in 2023 to make Google Search the default search engine on Apple devices. However, it's worth noting that this partnership won't be publicly acknowledged, with Apple planning to eventually develop its own model.
The company has reportedly been working on a one-trillion-parameter cloud-based model for consumer applications and aims to have it ready as early as next year. The integration of Gemini into Siri marks the latest move in Apple's pursuit of AI capabilities, following earlier rumors of partnerships with Anthropic and OpenAI.
While this development comes after Apple delayed Siri's release by a year in 2025, the need for enhanced assistance appears pressing. As reported in August last year, Google was asked to create a version of its Gemini model that could run securely on Apple's servers, highlighting the potential for AI integration into the virtual assistant.
According to Bloomberg, the Gemini model will play a critical role in handling Siri's summarizer and planner functions, which enable the virtual assistant to synthesize information and execute complex tasks. Given Siri's core function of using apps on behalf of users, the partnership with Google appears strategic.
Sources indicate that Apple is willing to pay $1 billion annually for access to Gemini technology, a significant chunk of change compared to the rumored $10 billion payment made by Google in 2023 to make Google Search the default search engine on Apple devices. However, it's worth noting that this partnership won't be publicly acknowledged, with Apple planning to eventually develop its own model.
The company has reportedly been working on a one-trillion-parameter cloud-based model for consumer applications and aims to have it ready as early as next year. The integration of Gemini into Siri marks the latest move in Apple's pursuit of AI capabilities, following earlier rumors of partnerships with Anthropic and OpenAI.
While this development comes after Apple delayed Siri's release by a year in 2025, the need for enhanced assistance appears pressing. As reported in August last year, Google was asked to create a version of its Gemini model that could run securely on Apple's servers, highlighting the potential for AI integration into the virtual assistant.